AIS Chapter 5 and 7

How is big data changing business and accounting?

Big data allows organizations to analyze massive amounts of information to improve decision-making, forecasting, auditing, tax planning, risk assessment, customer service, and operational efficiency.

Examples:

• Auditors test entire populations instead of samples.

• Tax professionals estimate tax impacts in real time.

• Airlines predict arrival times more accurately.

What are the benefits of big data?

• Better decision-making

• Improved forecasting

• Increased efficiency

• Better customer insights

• Competitive advantage

• Reduced costs

What are challenges of big data?

• Data quality issues

• Privacy concerns

• Data security risks

• Storage costs

• Complex analysis requirements

• Data becoming dark data or a data swamp

What are the 4 V's of big data?

1. Volume

2. Velocity

3. Variety

4. Veracity

What is Volume?

The amount of data collected and stored.

Example: Amazon storing billions of customer transactions.

What is Velocity?

The speed at which data are generated and processed.

Example: Credit card transactions occurring every second.

What is Variety?

Different forms of data.

Examples:

• Text

• Images

• Videos

• Emails

• Social media posts

What is Veracity?

The quality, accuracy, and trustworthiness of data.

Example: Removing duplicate customer records.

How can each V impact the data processing cycle?

V	Impact
Volume	More storage and processing needed
Velocity	Faster data processing required
Variety	More complex ETL process
Veracity	More validation and cleaning required

What are the four steps of the analytics mindset?

1. Ask the right question

2. Extract, Transform, Load (ETL)

3. Apply analytics

4. Interpret and share results

What makes a good data analytics question?

SMART

• Specific

• Measurable

• Achievable

• Relevant

• Timely

Example of a SMART analytics question?

"What factors increased sales revenue during Q1 2026?"

Specific, measurable, relevant, and time-bound.

What does ETL stand for?

Extract
Transform
Load

What are the three steps in data extraction?

Understand data needs

Extract data

Verify and document extraction

What is metadata?

Data about data.

Examples:

• Data type

• Field length

• Date format

What is a data dictionary?

Documentation describing the structure and meaning of data fields.

What is a primary key?

A field that uniquely identifies each record.

Example: Employee ID

What occurs during transformation?

• Standardizing

• Cleaning

• Structuring

• Formatting data

Why is data transformation often the most time-consuming step?

Data from different systems frequently have different formats and structures.

What happens during loading?

Transformed data are imported into a database, software application, or analysis tool.

What is structured data?

Highly organized data stored in fixed fields.

Examples:

• Relational databases

• Spreadsheets

What is unstructured data?

Data with no predefined format.

Examples:

• Images

• Videos

• Emails

• Social media posts

What is semi-structured data?

Data that has some organization but not enough for a relational database.

Examples:

• XML

• JSON

• CSV files

What is a data warehouse?

A repository of structured data from many sources.

What is a data mart?

A smaller subset of a data warehouse.

Example: Sales department data mart.

What is a data lake?

Storage for structured, semi-structured, and unstructured data.

What is dark data?

Data collected but never analyzed.

What is a data swamp?

Poorly documented data that cannot be effectively identified or analyzed.

How can companies avoid dark data and data swamps?

• Maintain accurate data dictionaries

• Document metadata

• Regularly analyze data

• Invest in data governance

What question does descriptive analytics answer?

"What happened?"

Examples of descriptive analytics?

• Profit margin

• Inventory turnover

• Current ratio

• Budget vs actual

What question does diagnostic analytics answer?

"Why did it happen?"

Example of diagnostic analytics?

Investigating why gross profit decreased.

What is the 5 Whys technique?

Continuously asking "Why?" to find the root cause.

What question does predictive analytics answer?

"What will likely happen?"

Examples of predictive analytics?

• Forecasting sales

• Predicting customer churn

• Fraud prediction

What question does prescriptive analytics answer?

"What should we do?"

Examples of prescriptive analytics?

• Loan approval systems

• Inventory reorder recommendations

• Route optimization

What is the mean?

Arithmetic average.

What is the median?

Middle value.

Why compare mean and median?

Large differences may indicate outliers.

What is an outlier?

An observation far from most other observations.

What measures spread?

• Range

• Standard deviation

• Quartiles

What is a correlation coefficient?

Measure of relationship strength between variables.

Range:
-1 to +1

Does correlation imply causation?

No.

Example of correlation without causation?

Snow gear sales and advertising both increase before winter. Winter may be causing both.

What is a null hypothesis (H₀)?

No relationship exists.

What is an alternative hypothesis (H₁)?

A relationship exists.

What is a Type I Error?

Rejecting a true null hypothesis.

False positive.

What is a Type II Error?

Failing to reject a false null hypothesis.

False negative.

Typical significance level?

0.05

Why visualize data?

• Easier to understand

• Faster processing

• Better communication

• Supports decision-making

What is data storytelling?

Translating analytics into understandable information for stakeholders.

Best chart for comparisons?

Bar chart

Best chart for showing progress toward goals?

Bullet chart

Best chart for correlation?

Scatterplot

Best chart for correlation using color intensity?

Heatmap

Best chart for distributions?

• Histogram

• Boxplot

Best chart for trends over time?

• Line chart

• Area chart

Best chart for part-to-whole relationships?

• Pie chart

• Treemap

What are the three major visualization design principles?

   Simplification

   Emphasis

Ethical presentation

Four simplification techniques?

Quantity

Distance

Orientation

Color

What is quantity?

Using only the necessary amount of information.

What is distance?

Keeping related information close together.

What is orientation?

Presenting information horizontally when possible and sorting logically.

Four emphasis techniques?

1. Highlighting

2. Weighting

3. Ordering

4. Color

Examples of highlighting?

• Arrows

• Labels

• Contrast

• Color

What increases visual weight?

• Larger size

• Darker color

• Greater contrast

• More density

Why is ordering important?

It helps emphasize important values and improves interpretation.

What is confirmation bias?

Interpreting information in a way that supports existing beliefs.

How can visualizations misrepresent data?

• Truncated axes

• Misleading scales

• Distorted proportions

• Excessive emphasis

How do you avoid data misrepresentation?

• Use honest scales

• Label clearly

• Present complete information

• Avoid deceptive formatting

• Maintain objectivity

An auditor calculates inventory turnover and profit margin. What type of analytics?

Descriptive

A company investigates why profits decreased. What type?

Diagnostic

A bank predicts whether a customer will default on a loan. What type?

Predictive

A system automatically approves or rejects a loan application. What type?

Prescriptive

A company stores emails, videos, and transaction data together. What storage structure?

Data Lake

Which visualization would best show sales trends over five years?

Line Chart

Which visualization would best show relationship between training hours and performance?

Scatterplot

Which visualization would best show percentage of expenses by category?

Pie Chart or Treemap