Chapter 1: Ask the Question: Using Data Analytics to Address Accounting Questions

Data growth

Data is increasing exponentially; the slides state that the data flooding the Internet every second is equivalent to the data stored on the entire Internet 20 years ago.

Accounting's information role

Accounting bills itself as the critical information provider; accountants need to harness data and machines to provide better information to decision makers.

Accounting knowledge and automation

Accounting knowledge informs automation.

Data analytics and accounting expertise

Data analytics is one way for accountants to develop and exhibit expertise needed in an increasingly computerized environment.

Computers vs. accountants

Computers increasingly collect data, but they do not have the accounting expertise needed; accountants provide expertise that automation does not provide.

Bloom's Taxonomy

A hierarchical view of critical thinking skills.

Three lower levels of Bloom's Taxonomy

Remember, understand, and apply.

Three higher levels of Bloom's Taxonomy

Analyze, evaluate, and create.

Machines and Bloom's Taxonomy

Machines excel in the three lower levels: remember, understand, and apply.

Accountants and Bloom's Taxonomy

Accountants must play a major role in the higher levels: analyze, evaluate, and create.

Remember

Recognizing and recalling facts.

Understand

Understanding what facts mean.

Apply

Applying facts, rules, concepts, and ideas.

Analyze

Breaking down information into component parts.

Evaluate

Judging the value of information and ideas.

Create

Combining parts to make a new whole.

Why basic knowledge matters

Accountants cannot analyze, evaluate, and create if they do not already have the basic accounting knowledge and understanding required for remembering, understanding, and applying.

AMPS model

A process for evaluating data analytics: Ask the Question, Master the Data, Perform the Analysis, Share the Story.

AMPS — A

Ask the Question.

AMPS — M

Master the Data.

AMPS — P

Perform the Analysis.

AMPS — S

Share the Story.

Ask the Question

The first AMPS step: identify an accounting question that can potentially be answered with data.

Example question: profitability by location

Which product is more profitable at stores in Arizona?

Example question: production location

Is it more profitable to produce an item in the U.S. or in the Philippines?

Example question: overhead

How much overhead should be applied to each of five Android apps?

Example question: investment risk

What is the probability that a company being considered for investment will go bankrupt or commit financial statement fraud?

Example question: breakeven

As fixed and variable costs change, how will that affect the breakeven point in sales?

Example question: allowance

What is a reasonable estimate for the allowance for doubtful accounts based on past experience?

Descriptive analytics

Answers: What happened? What is happening?

Diagnostic analytics

Answers: Why did it happen? What are the root causes of past results?

Predictive analytics

Answers: Will it happen in the future? What is the probability something will happen? Is it forecastable?

Prescriptive analytics

Answers: What should we do based on what we expect will happen? How can we optimize performance based on potential constraints?

Master the Data — core questions

What issues exist in the data? Can the data answer/address the question?

Data integrity

Data is accurate, valid, and consistent.

Data-quality checks

Determine whether data has errors, is missing, or is biased.

Data types

Categorical versus numerical.

Data analysis capability

Consider what type of analysis the data allows you to do.

Cost-benefit of data

Consider the cost of acquiring data versus the benefit of using the data.

Data ownership/access

Consider who owns the data and whether the data is hard to access.

Perform the Analysis

Use appropriate models and analytical techniques to gain insight from the data.

PivotTable use

Can summarize accounts receivable aging or post journal entries to a trial balance.

Cash flow analysis

Can include internal rate of return (IRR) or net present value (NPV) to evaluate capital investments or equity investments.

Histogram/scatterplot in internal controls

Can help evaluate journal entries that are excessively big, excessively small, or negative when testing internal controls.

Regression analysis

Can evaluate cost behavior by segregating total costs into fixed and variable cost components.

Identifying shifts and outliers

Informs decision-making.

What-If/Goal Seek

Can analyze how changing costs and other factors affect the breakeven level for a new product.

Benford's law

Can be used to find and evaluate unusual and potentially fraudulent transactions.

Share the Story

The AMPS step focused on choosing the best way to communicate what was found in data analysis.

Static visualization

A one-off analysis.

Examples of static outputs

Reports, graphs, and tables.

Dynamic visualization

A visualization that updates on a continuous basis.

Example of a dynamic visualization

A dashboard.

Daily sales vs. sales targets

A dynamic report would be appropriate because the data needs to be tracked and updated continuously.

Purpose: comparison of values

Example: Did Toyota or Tesla have greater sales last year? Common visualization: bar chart.

Purpose: composition of values

Example: What proportion of sales comes from each division? Common visualization: pie chart.

Purpose: distribution of values

Example: How many days did it take to fulfill a recent set of 20,000 orders? Common visualization: histogram.

Purpose: trends over time

Example: How has Google's ad revenue changed over the last five years? Common visualization: line graph.

Purpose: relationships between values

Example: What is the relationship between advertising expense and sales revenue for Costco? Common visualization: scatterplot.

Bar chart

Common visualization for comparison of values.

Pie chart

Common visualization for composition of values.

Histogram

Common visualization for distribution of values.

Line graph

Common visualization for trends of values over time.

Scatterplot

Common visualization for relationships between values.

Master the Data tools

Excel (basic), Alteryx, SQL, Tableau Prep, and Power Query in Power BI.

SQL

Structured Query Language; used to access specific data from very large datasets.

Alteryx

A tool used to help prepare data for analysis.

Tableau Prep

A tool used to help prepare data for analysis.

Power Query

The data preparation capability associated with Power BI listed in the slides.

Excel PivotTables

A tool that allows reorganization and summarization of certain data using crosstabulations without changing the underlying spreadsheet or data.

Excel Data Analysis ToolPak

Provides access to specific analysis techniques including descriptive statistics, histograms, correlation, and regression.

Interactive data analysis tools

Excel basic analysis, including PivotTables and the Data Analysis ToolPak.

Programmatic analysis tools

SAS, SPSS, Stata, R, and Python.

SAS, SPSS, and Stata

Software analysis tools that emphasize statistics.

R and Python

Programming software tools used to perform advanced data analysis.

GenAI tools

ChatGPT, Gemini, Claude, Copilot, and other generative AI tools can perform analyses, create visualizations, and explain relationships in data when given appropriate guidance and natural-language prompts.

Visualization tools

Excel, Tableau Desktop, Power BI, and GenAI.

Tableau

Used for advanced data visualization in the slides.

Power BI

Used for advanced data visualization and includes Power Query for data preparation.

GenAI definition

A type of artificial intelligence that analyzes large data sets and generates new audio, images, music, text, videos, and other content in response to user prompts.

GenAI natural-language capability

GenAI tools make it easy to describe in natural language the results or analyses desired, and they generally produce the output as instructed.

Lab 1-1

Excel — Journal Entries to Trial Balance.

Lab 1-2

Excel — Using SLN(), SYD(), and DDB() functions to depreciate fixed assets.

Lab 1-3

Excel — Creating a Mortgage Amortization Schedule.

Lab 1-4

Excel — Creating an Adjusted Trial Balance Using a Pivot Table.

Chapter summary: computerization

Because of computerization and the mounds of data available, accountants will need to be adept at data analytics.

Chapter summary: AMPS

The AMPS model is the foundation for the rest of the book and highlights the need to ask appropriate questions that are potentially answerable via data analyses.

Chapter summary: accountant capabilities

Accountants need new capabilities, including higher-order thinking, new skills, and new tools.

Fill in the blank: AMPS

The four AMPS steps are Ask the Question, Master the Data, Perform the Analysis, and Share the Story.

Fill in the blank: descriptive analytics

_____ analytics asks what happened or what is happening.

Fill in the blank: diagnostic analytics

_____ analytics asks why something happened and seeks root causes.

Fill in the blank: predictive analytics

_____ analytics asks whether something will happen in the future and its probability.

Fill in the blank: prescriptive analytics

_____ analytics asks what should be done based on expected outcomes and constraints.

Fill in the blank: visualization trend

A _____ graph is commonly used to show trends over time.

Fill in the blank: relationship

A _____ is commonly used to visualize relationships between values.

Fill in the blank: distribution

A _____ is commonly used to visualize the distribution of values.

Fill in the blank: comparison

A _____ chart is commonly used to compare values.

Fill in the blank: composition

A _____ chart is commonly used to show the composition of values.

Question: why ask the right question?

Because data will not provide useful insight unless the accountant asks the right data analytics questions.

Question: what should be checked before analysis?

Check whether the data can answer the question, has integrity, contains errors or missing values, is biased, permits the desired analysis, is worth its acquisition cost, and is accessible/owned appropriately.