Data Analytics in Accounting: Big Data, ETL, and Visualization

What is big data?

Describes a massive amount of data captured, stored, and analyzed.

What are the four V's of big data?

Volume, Velocity, Variety, Veracity.

What does Volume mean in big data?

Amount of data.

What does Velocity mean in big data?

Speed of data.

What does Variety mean in big data?

Different forms data can take.

What does Veracity mean in big data?

Quality and trustworthiness of data.

What does it mean to have an analytics mindset?

Using data to answer questions and make decisions.

What are SMART objectives?

Specific, Measurable, Achievable, Relevant, Timely.

What is the ETL process?

Extract, Transform, Load.

What happens in the extract phase of ETL?

Understand data needs, perform extraction, and verify quality.

What happens in the transform phase of ETL?

Standardize structure, clean data, validate quality, document transformation.

What happens in the load phase of ETL?

Store data in a compatible format and create a data dictionary.

What is structured data?

Highly organized data in tables.

What is semi-structured data?

Data not organized enough for databases but has some structure, like Excel or CSV files.

What is unstructured data?

Data like text and images that is not organized.

What is a flat file?

A file that contains all data together.

What is a delimiter?

A character that separates fields, usually |.

What is a text qualifier?

A symbol that indicates the beginning and end of a field.

What are the four steps in the data transformation process?

1. Understand data and desired outcome, 2. Standardize and clean data, 3. Verify quality, 4. Document the process.

What is descriptive analysis?

Analysis that answers 'What happened?' using historical data.

What is diagnostic analysis?

Analysis that answers 'Why did that happen?' to find trends.

What is predictive analysis?

Analysis that answers 'What might happen?' using regression models.

What is prescriptive analysis?

Analysis that answers 'What should we do?' and provides recommended actions.

What is data storytelling?

Translating complex data analysis into easily understandable terms.

What is a data visualization?

Graphical representation to convey meaning.

What is a data dashboard?

Visual display of important statistics.

What is robotic process automation (RPA)?

Programmed tasks across applications, used for ETL.

What are the attributes of high-quality data?

Accurate, complete, current, consistent, timely, valid.

What is data structuring?

Putting data in a usable format.

What is aggregate data?

Summarized data.

What is data joining?

Combining multiple sets of data.

What is data pivoting?

Flipping rows and columns.

What is data standardization?

Putting data in a common format.

What is data parsing?

Breaking data apart.

What is data concatenation?

Putting data together.

What are cryptic data values?

Values that require a code to understand.

What is a dummy or dichotomous variable?

A variable with only two options, usually yes/no or 0/1.

What are misfielded data values?

Correctly formatted data in the wrong field.

What is data consistency?

Every value in a field is stored the same.

What is dirty data?

Data that is not consistent, accurate, or complete.

What is data cleaning?

Updating data to be consistent, accurate, and complete.

What is data deduplication?

Removing duplicate data.

What is data filtering?

Removing unnecessary data for analysis.

What is data imputation?

Replacing null/missing data with a substituted value.

What are data contradiction errors?

Conflicting descriptions of the same entity.

What are data threshold violations?

Data that is too big or small, outside allowable limits.

What are violated attribute dependencies?

An attribute that doesn't match the primary key.

What are data entry errors?

Errors caused by incorrect data entry.

What is data validation?

Analyzing data to ensure high quality before, during, and after transformation.

What is exploratory data analysis?

Used for descriptive analysis to let data tell the story.

What is an outlier?

A value that lies an abnormal distance from others.

What is confirmatory data analysis?

Analysis used to test hypotheses.

What is a null hypothesis?

A hypothesis stating there is no difference.

What is an alternative hypothesis?

A hypothesis stating there is a difference.

What is a type I error?

Rejecting a true null hypothesis.

What is a type II error?

Accepting a false null hypothesis.

What is categorical data?

Attribute data or qualitative data.

What type of visualization is good for comparison?

Bar chart or bullet chart.

What type of visualization is good for correlation?

Scatterplot or heatmap.

What type of visualization is good for distribution?

Histogram or box plot.

What type of visualization is good for trend evaluation?

Line chart or area chart.

What type of visualization is good for part of a whole?

Tree chart or pie chart.

What does simplification in visualization refer to?

Making visualizations easy to understand.

What does emphasis in design mean?

Making the most important message easily identifiable.

What does ethical data presentation refer to?

Avoiding deception that changes understanding.

How can we simplify visualizations?

Show all needed info, use data labels, minimize distractions.

How can we emphasize data in visualizations?

Use color, size, and contrast.

How can we ethically present data in visualizations?

Start y-axis at 0, show full graph, avoid bad weighting.