Transcript Board Monitoring and Earnings Management: Do Outside

Data Mining As A Continuous Auditing Tool
for “Soft Information”: A Research Question
A Research Proposal
By
J. Donald Warren, Jr.
Rutgers University
Fifth Continuous Assurance Symposium
Rutgers Business School
November 22-23, 2002
Data Mining As a Continuous Auditing Tool for “Soft
Information”: A Research Question
What this proposal is about
• Auditors will have to address “soft information”
in a continuous auditing environment.
• “Soft information” defined as management
estimates and/or judgments
• Timing of traditional audit procedures may not
be appropriate in continuous audit environment
• Data mining may provide necessary tools to
provide
competent
evidence
for
“soft
information”
in
a
continuous
auditing
environment
Data Mining As a Continuous Auditing Tool for “Soft
Information”: A Research Question
How this proposal contributes to body of knowledge
• A continuous audit will require tools and
techniques to provide the auditor with
sufficient competent evidence on a “real time”
basis.
• Data mining has been used in other disciplines
with success for profiling customers or in
forensic projects
• Valuation accounts in a continuous auditing
environment creates challenges to auditors in a
continuous auditing environment
Data Mining As a Continuous Auditing Tool for “Soft
Information”: A Research Question
Development of Research Question
• Allowance for doubtful accounts selected
as “soft information” account
• Traditionally allowance for doubtful
accounts has been the subject of
differing opinions between auditors and
clients
• Auditors
cannot
apply
traditional
procedures in a continuous auditing
environment
Data Mining As a Continuous Auditing Tool for “Soft
Information”: A Research Question
Development of Research Question (continued)
• Embedded audit modules may be used to
address “soft information” accounts
• In case of allowance for doubtful
accounts, auditor will be unable to
discuss questionable accounts and review
subsequent collections for accounts
receivables at the date that a “soft
information” account may be reported
upon
Data Mining As a Continuous Auditing Tool for “Soft
Information”: A Research Question
Development of Research Question (continued)
• Data mining provides a process to ferret
through voluminous data to identify
patterns
• Research question:
• In a continuous auditing environment, do data
mining techniques provide auditors with a
methodology which generates sufficient competent
evidential matter on accounts containing “soft
information” and enable auditors to rely on the
results computed as being adequate for such
accounts?
Data Mining As a Continuous Auditing Tool for “Soft
Information”: A Research Question
Research Methodology
• Obtain a database for a company
which contains accounts receivables
as well as other accounts which
may be necessary to test for the
adequacy of such accounts in a
continuous auditing enviroment
Data Mining As a Continuous Auditing Tool for “Soft
Information”: A Research Question
Research Methodology
• Use data mining tools such as
stepwise regressions, decision trees
and logistic regressions to address
“soft information”
• Three samples required: training,
testing and validation