ACC 490 Final

What is the SPARKS framework?

A critical thinking toolkit

What are the four types of data analysis methods?

Descriptive, Diagnostic, Predictive, Prescriptive

What is the objective of descriptive data analytics?

To investigate what is happening currently or has occurred in the past

What is the objective of diagnostic analytics?

To help understand why something happened

What is the objective of predictive analytics?

To forecast what might happen in the future

What is the objective of prescriptive analytics?

To help understand what should happen to meet goals and objectives

What are some common data analysis techniques for descriptive analytics?

Sum, count, average, median, standard deviation

What are some common data analysis techniques for diagnostic analytics?

Anomaly and outlier detection, trend analysis, and pattern recognition

What are some common data analysis techniques for predictive analytics?

Forecasting, regression analysis, and time-series analysis

What are some common data analysis techniques for prescriptive analytics? (duplicate)

Optimization and what-if analyses (duplicate)

What is the data analytics mindset?

The professional habit of critically thinking throughout the data analysis process before making and communicating a professional choice or decision

What skills are required for the data analytics mindset?

Critical thinking, data literacy, technological agility, communication skills

What are the 6 components of the SPARKS framework?

Stakeholders, Purpose, Alternatives, Risks, Knowledge, Self-Reflection

What are potential risks of data analysis?

Inappropriate/incomplete data, inappropriate or incorrectly applied method, not understanding or evaluating assumptions about the data or results, having biases that could influence the data analysis

What is an attribute?

It is the data fields that describe different aspect of the records (column)

What is the difference between a null value and zero?

A null value indicates there is no data present, a zero means that the data is zero. They do not mean the same thing

How can descriptive statistics be used to understand the characteristics of data?

To see the average observations in the data, the data's shape, and the distribution of the data

What measures can be used to describe the data's shape and distribution?

Skewness and Kurtosis

What type of visualizations are used to show composition?

Area chart, pie chart, stacked bar chart

What type of visualizations are used to show relationships?

Bubble chart, scatter chart

What type of visualizations are used to show distributions?

Histogram chart, line chart, scatter chart

What type of visualizations are used to indicate trends?

Line chart, column chart

What type of visualizations are used to show comparisons?

Bar chart, column chart, line chart

What questions should be asked for descriptive analytics?

Questions that identify the purpose of the analysis

What types of analysis can be performed to answer descriptive questions?

Frequency measures, measures of location, measures of dispersion, measures of percentage change

What questions should be asked for diagnostic analytics?

Questions that explore the data to find the cause of an outcome and look for anomalies, correlations, patterns, and trends

What types of analysis can be performed to answer diagnostic questions?

Anomaly detection, correlation, pattern detection, and trend analysis

What questions should be asked for predictive analytics?

Questions that identify trends and focus on what may happen in the future

What types of analysis can be performed to answer predictive questions?

Trendlines, linear regression

What questions should be asked for prescriptive analytics?

Questions that focus on what should happen or may investigate how to take advantage of future opportunities or mitigate future risk outcomes

What types of analysis can be performed to answer prescriptive questions?

Linear optimization, What-If analyses

What is data preparation?

The process of profiling, cleaning, restructuring, and integrating data prior to processing analysis

What steps does data preparation entail?

Profiling, cleaning, restructuring, and integrating

What are flat tables?

Tables where column titles do not contain data useful for analysis

What are crosstab tables?

Tables where column titles contain data useful for analysis

What are composite tables?

A table that combines values for two or more characteristics

What is the recommended structure for analytical data models?

Star schema

Why is star scheme the recommended structure for analytical data models?

It is the recommended structure because it is easy to understand by business users, and computers can process them efficiently

What is the ETL process?

Extract, transform, and load, it is the process of correcting issues with data

What does the extract step of the ETL process entail?

Source data is moved to the platform where they will be transformed, and data validation or confirming that the data were transferred completely and correctly

What does the transform step of the ETL process entail?

It improves the raw data for analysis through cleaning, restructuring, and integration

What does the load step of the ETL process entail?

Making the analytical database available for use

What is information modeling?

The process of generating additional knowledge from data that is relevant for analysis purposes

What are calculated columns?

A new field created from an in-table numeric calculation, each record has a value in it

What are measures?

An aggregate, or total, that can be used in reports, and thus for analytical purposes, they are created by algorithms, but are not integral parts of a table

What are the various data relationships?

Nominal comparison, distribution, deviation, ranking, part-to-whole, correlation, time series, geospatial

What charts/graphs can be used to visualize geospatial relationships?

Maps like choropleth (color intensity) and proportional symbol (bigger bubbles for bigger data values)

What charts/graphs can be used to visualize time series relationships?

Line chart, but also bar, column, area, waterfall, and sparkline charts

What charts/graphs can be used to visualize correlations?

Scatterplot/Scatter chart

What charts/graphs can be used to visualize part-to-whole relationships?

Pie charts, donut charts, stacked bar charts, stacked column charts, and treemaps

What charts/graphs can be used to visualize ranking relationships?

None specific, but many things like table, bar charts, and column integrate ranking information

What charts/graphs can be used to visualize deviation relationships?

Clustered bar charts, clustered column charts, gauges, bullet charts

What charts/graphs can be used to visualize distributions?

Histograms, violin plots, and box-and-whisker charts

What charts/graphs can be used to visualize nominal comparisons?

Bar charts, column charts, dot plots, and lollipop charts

What are the four steps to data exploration?

Identify questions, identify data relationships, explore data relationships, generate insights

What are the two types of data exploration patterns?

Composite trends & Pareto analysis

What are visualizations used for composite trends?

Stacked Area Chart, Stacked Column Chart

What are visualizations used for Pareto analysis?

Pareto chart, or line and column charts.

What is data analysis interpretation?

The process of evaluating an analysis to understand and explain its meaning.

What is the detailed consideration of the stakeholders part of the SPARKS framework?

Identify the stakeholders (internal and external) of the decision facing your team

What is the detailed consideration of the purpose part of the SPARKS framework?

Identify the purpose of the analysis for effective interpretation

What is the detailed consideration of the alternatives part of the SPARKS framework?

Consider are there alternative ways to view the results of the analysis or to conduct the analysis that were not addressed

What is the detailed consideration of the risks part of the SPARKS framework?

Consider data risks, is the data complete, accurate, composed of timely/relevant data, and are appropriate controls in place to ensure the data was used correctly. Also consider analysis risks: whether the correct method was used, whether the analysis used the wrong data, and whether the analysis answered the purpose/question. Lastly, consider potential biases.

What is the detailed consideration of the knowledge part of the SPARKS framework?

What knowledge is required to interpret the analysis: specific accounting, industry, technology.

What is the detailed consideration of the self-reflection part of the SPARKS framework?

Think about lessons learned and how past experience can help evaluate the current analysis and the current analysis can help with future ones

What are the different types of biases?

Data biases, preparer biases, evaluator biases, confirmation bias, and selection bias.

What are data biases?

Were all the necessary and appropriate data included in the analysis

What are preparer biases?

Did the preparer have any potential biases that could have influenced the preparation of the analysis?

What are evaluator biases?

Does the evaluator (you) have any biases that could influence your interpretation of the results

What is confirmation bias?

Performing the analysis to prove a predetermined assumption, looking for data or results to confirm existing belief

What is selection bias?

Bias demonstrated in the subjective selection of data used in the analysis

What are the different data analytics purpose categories? (duplicate)

Descriptive, Diagnostic, Predictive, and Prescriptive (duplicate)

What type of analyses are performed under the descriptive analytics category? (duplicate)

Frequency distribution, cross tabulation, measures of location, measures of dispersion (duplicate)

What type of analyses are performed under the diagnostic analytics category? (duplicate)

Anomaly detection, correlation analysis, trend analysis (duplicate)

What type of analyses are performed under the predictive analytics category?

Linear regression and predictive algorithms

What type of analyses are performed under the prescriptive analytics category?

Optimization models and what-if analyses

What is data exploration?

The process of analyzing the data to determine whether additional analyses is required.