USING COMPUTER
ASSISTED AUDIT
TOOLS AND
TECHNIQUES
(CAATT’S)
Computer Assisted Audit Tools and Techniques
• Has two subcomponents:
1. Software used to increase an auditor’s
personal productivity and software used
to perform data extraction and analysis,
2. Techniques to increase the efficiency
and effectiveness of the audit function.
Input Controls
• Designed to ensure that the
transactions that bring data into
the system are valid, accurate, and
complete
Data input procedures can be
either:
Source document-triggered (batch)
Direct input (real-time)
Source document input requires
human involvement and is prone to
clerical errors.
Direct input employs real-time
editing techniques to identify and
correct errors immediately
Classes of Input Controls
1) Source document controls
2) Data coding controls
3) Batch controls
4) Validation controls
5) Input error correction
6) Generalized data input systems
Source Document Controls
Controls in systems using physical source
documents
To control for exposure, control
procedures are needed over source
documents to account for each one
Use pre-numbered source documents
Use source documents in sequence
Periodically audit source documents
Data Coding Controls
Checks on data integrity during processing
Transcription errors
Addition errors
Truncation errors
Substitution errors
Transposition errors
Single transposition
Multiple transposition
Control = Check digits
Added to code when created (suffix, prefix,
embedded)
Sum of digits (ones): transcription errors only
Modulus 11: different weights per column:
transposition and transcription errors
Introduces storage and processing inefficiencies
Batch Controls
Method for handling high volumes of
transaction data – esp. paper-fed IS
Controls of batch continues thru all phases
of system and all processes (i.e., not JUST an
input control)
1) All records in the batch are processed together
2) No records are processed more than once
3) An audit trail is maintained from input to output
Requires grouping of similar input
transactions
Batch Controls
Requires controlling batch throughout
Batch transmittal sheet (batch control record)
Unique batch number (serial #)
A batch date
A transaction code
Number of records in the batch
Total dollar value of financial field
Sum of unique non-financial field
• Hash total
• E.g., customer number
Batch control log
Hash totals
Validation Controls
Intended to detect errors in data
before processing
Most effective if performed close to
the source of the transaction
Some require referencing a master
file
Field Interrogation
Missing data checks
Numeric-alphabetic data checks
Zero-value checks
Limit checks
Range checks
Validity checks
Check digit
Record Interrogation
Reasonableness checks
Sign checks
Sequence checks
File Interrogation
Internal label checks (tape)
Version checks
Expiration date check
Input Error Connection
Batch – correct and resubmit
Controls to make sure errors dealt
with completely and accurately
1) Immediate Correction
2) Create an Error File
Reverse the effects of partially
processed, resubmit corrected
records
Reinsert corrected records in
processing stage where error was
detected
3) Reject the Entire Batch
Generalized Data Input Systems (GDIS)
Centralized procedures to manage data
input for all transaction processing systems
Eliminates need to create redundant
routines for each new application
Advantages:
Improves control by having one
common system perform all data
validation
Ensures each AIS application applies a
consistent standard of data validation
Improves systems development
efficiency
Major components:
1) Generalized Validation
Module
2) Validated Data File
3) Error File
4) Error Reports
5) Transaction Log
Process Controls
1) Run-to-Run Controls
2) Operator Intervention Controls
3) Audit Trail Controls
Run-to-Run (Batch)
Use batch figures to
monitor the batch as it
moves from one process
to another
1) Recalculate Control
Totals
2) Check Transaction Codes
3) Sequence Checks
Operator Intervention
When operator manually enters
controls into the system
Preference is to derive by logic or
provided by system
Audit Trail Controls
Every transaction becomes
traceable from input to output
Each processing step is documented
Preservation is key to auditability of
AIS
Transaction logs
Log of automatic transactions
Listing of automatic transactions
Unique transaction identifiers [s/n]
Error listing
Output Controls
Ensure system output:
1) Not misplaced
2) Not misdirected
3) Not corrupted
4) Privacy policy not violated
Batch systems more susceptible to exposure,
require greater controls
Controlling Batch Systems Output
Many steps from printer to end user
Data control clerk check point
Unacceptable printing should be shredded
Cost/benefit basis for controls
Sensitivity of data drives levels of controls
Output spooling – risks:
Access the output file and change
critical data values
Access the file and change the
number of copies to be printed
Make a copy of the output file so
illegal output can be generated
Destroy the output file before
printing take place
Print Programs
Operator Intervention:
1) Pausing the print program to load output paper
2) Entering parameters needed by the print run
3) Restarting the print run at a prescribed checkpoint after
a printer malfunction
4) Removing printer output from the printer for review and
distribution
Print Program Controls
Production of unauthorized copies
Employ output document controls similar to source
document controls
Unauthorized browsing of sensitive data by
employees
Special multi-part paper that blocks certain fields
Bursting
Supervision
Waste
Proper disposal of aborted
copies and carbon copies
Data control
Data control group – verify
and log
Report distribution
Supervision
End user controls
End user detection
Report retention:
Statutory requirements (gov’t)
Number of copies in existence
Existence of softcopies (backups)
Destroyed in a manner consistent
with the sensitivity of its
contents
Controlling real-time systems
output
Eliminates intermediaries
Threats:
Interception
Disruption
Destruction
Corruption
Exposures:
Equipment failure
Subversive acts
Systems performance controls
Chain of custody controls
TESTING COMPUTER
APPLICATION CONTROLS
1) Black box (around)
2) White box (through)
Black Box Testing
Ignore internal logic of application
Use functional characteristics
Flowcharts
Interview key personnel
Advantages:
Do not have to remove application from
operations to test it
Appropriately applied:
Simple applications
Relative low level of risk
White Box Testing
Relies on in-depth understanding of
the internal logic of the application
Uses small volume of carefully
crafted, custom test transactions to
verify specific aspects of logic and
controls
Allows auditors to conduct precise
test with known outcomes, which
can be compared objectively to
actual results
White Box Tests Methods
1) Authenticity tests:
Individuals / users
Programmed procedure
Messages to access system (e.g., logons)
All-American University, student lab:
logon, reboot, logon *
2) Accuracy tests:
System only processes data values that
conform to specified tolerances
3) Completeness tests:
Identify missing data (field, records,
files)
4) Redundancy tests:
Process each record exactly once
5) Audit trail tests:
Ensure application and/or system
creates an adequate audit trail
Transactions listing
Error files or reports for all exceptions
6) Rounding error tests:
“Salami slicing”
Monitor activities – excessive ones are
serious exceptions; e.g, rounding and
thousands of entries into a single
account for $1 or 1¢
Computer Aided Audit Tools and
Controls (CAATTs)
1) Test data method
2) Base case system evaluation
3) Tracing
4) Integrated Test Facility [ITF]
5) Parallel simulation
6) GAS
Test Data Method
Used to establish the application
processing integrity
Uses a “test deck”
Valid data
Purposefully selected invalid data
Every possible:
Input error
Logical processes
Irregularity
Procedures:
1) Predetermined results and expectations
2) Run test deck
3) Compare
Base Case System Evaluation
Variant of Test Data method
Comprehensive test data
Repetitive testing throughout SDLC
When application is modified,
subsequent test (new) results can
be compared with previous results
(base)
Tracing
Test data technique that takes step-
by-step walk through application
1) The trace option must be enabled for
the application
2) Specific data or types of transactions are
created as test data
3) Test data is “traced” through all
processing steps of the application, and
a listing is produced of all lines of code
as executed (variables, results, etc.)
Excellent means of debugging a
faculty program
Test Data: Pro’s and Cons
Pro’s Cons
– They employ white – Auditors must rely
box approach, thus on IS personnel to
providing explicit obtain a copy of the
application for
evidence
testing
– Audit evidence is not
– Can be employed entirely independent
with minimal – Provides static
disruption to picture of
operations application integrity
– Relatively high cost
– They require to implement,
auditing inefficiency
minimal computer
expertise on the
part of the auditors
Integrated Test Facility
ITF is an automated technique that
allows auditors to test logic and
controls during normal operations
Set up a dummy entity within the
application system
1) Set up a dummy entity within the
application system
2) System able to discriminate between ITF
audit module transactions and routine
transactions
3) Auditor analyzes ITF results against
expected results
Parallel Simulation
Auditor writes or obtains a copy of
the program that simulates key
features or processes to be
reviewed / tested
1) Auditor gains a thorough
understanding of the application under
review
2) Auditor identifies those processes and
controls critical to the application
3) Auditor creates the simulation using
program or Generalized Audit Software
(GAS)
4) Auditor runs the simulated program
using selected data and files
5) Auditor evaluates results and reconciles
differences
