150 likes | 165 Views
Data Analytics Conference. June 2019. bios. Highest risk and greatest value considerations when quantifying risk. What the data tells us using the 4 types of analytics to tell a story. Descriptive Analytics what happened in the past Diagnostic Analytics why something happened
E N D
Data Analytics Conference June 2019
Highest risk and greatest valueconsiderations when quantifying risk
What the data tells ususing the 4 types of analytics to tell a story • Descriptive Analytics • what happened in the past • Diagnostic Analytics • why something happened • Predictive Analytics • predicts the future based on the past • Prescriptive Analytics • determine what to do to mitigate a risk
What data do I needtaking what you have and identifying where you want to dig in deeper • Scientific Method • make an observation that describes a problem • create a hypothesis and key controls • test the hypothesis and key controls • draw conclusions and refine • Build data requirements on what data you need in order to test a hypothesis
Know your data Demo
Understanding the processwhat are the inputs, process and outputs? • Having a solid process map and understanding the flow of data will increase the value of your analysis • Questions to Consider: • How is the data transformed throughout the process? What rule exist in the process? What’s automated vs manual in the process?
Data sourcing strategygetting the data you want Tips and Tricks to help get the data you want: Have your hypothesis and testing method well defined – use this as your business case Build relationships with data owners throughout the company prior to needing the data Build trust by knowing the dos and don’ts of pulling data Be aware of any standard data access requests processes and start the conversation as early as possible
Data accuracy and completenessgetting the data you want • Understand parameters used to pull the data • Date Ranges • What is filtered out of data? (WHERE statements) • How is the data aggregated? Does it make sense? • Compare to known or related reports • Confirm data types are consistent throughout the report • Size of the data • Look at min/max/count/length • Nulls/Blanks • Leading and trailing whitespaces • Duplicates/Unique values
Hands on workshop Auditing Purchase Cards
Workshop agenda • Review of P_Card_Import Data : 6 years Data (2010-2015) • 1. What risks/controls are associated with P-Cards? • A. Spending Controls –Cardholder Spending Limits (Single cardholder –not to exceed $5000), Merchant Type activities – MCC allowed • B. Split Purchases • C. State Tax does not get charged • D. Gifts, Gift Cards, Gift Certificates are not allowed • E. No Gasoline purchase • F. Service awards – not allowed • G. Moving Expenses –now allowed • 2. Forming and testing controls • 3. Completeness and Accuracy Check • 4. Quick Analysis using Excel & Pivot Tables • 5. Conclusions
Checking completeness and accuracy • What data is contained within each record? • Compare to known or related reports • Confirm data types are consistent throughout the report • Size of the data • Look at min/max/count/length • Nulls/Blanks • Leading and trailing whitespaces • Duplicates/Unique values