Free DA0-002 Practice Questions – 2026 Updated

The question asks for the next logical step after the technical requirements for a reporting deck have been met. In a standard data visualization or business intelligence project lifecycle, after the initial development is complete, the solution must be validated with stakeholders. Showing a mock-up or a functional prototype to the team serves as a crucial review and user acceptance testing (UAT) step. This process ensures the visualizations are clear, the metrics are correct, and the security functions as intended before the final analysis is formally presented or the report is deployed to a wider audience. This iterative feedback loop is essential for project success.

B. Explain the desired level of reporting detail.

This is a requirements-gathering activity that must be completed before the analyst begins building the report.

C. Present an analysis of the data.

This is the final step of the project. It should only occur after the reporting deck has been reviewed and approved by the team or key stakeholders.

D. Find out the project due date.

This is a fundamental project management task that is established during the initial planning phase, long before the development work is completed.

---

1. Wirth

R.

& Hipp

J. (2000). CRISP-DM: Towards a standard process model for data mining. Proceedings of the 4th International Conference on the Practical Applications of Knowledge Discovery and Data Mining. Page 35

Section 3.5

"Evaluation". The CRISP-DM methodology

a standard in data projects

specifies an "Evaluation" phase that occurs after model/solution building but before final deployment. This phase involves reviewing the work against business objectives

which aligns with showing a prototype to the team for feedback.

2. Microsoft Corporation. (2023). Power BI adoption roadmap: BI solution planning. Microsoft Learn documentation. In the section "Project planning

" the documentation outlines the typical phases of a BI project

which include prototyping

development

and validation. The validation phase

which includes User Acceptance Testing (UAT)

occurs after development is complete and is essential before deployment. Showing a functional version to the team is a core part of this validation.

3. Munzner

T. (2014). Visualization Analysis and Design. CRC Press. Chapter 3

"A Nested Model for Visualization Design and Validation

" outlines a structured process for creating visualizations. The model emphasizes validation at each stage. After the "Algorithm Design" stage (equivalent to building the report)

the process requires validation to ensure the implementation is correct and effective before it is deployed for use. This academic model supports the practice of internal review and feedback post-development.

A data warehouse is the most appropriate choice. It is a centralized repository designed to integrate structured data from various sources across an entire enterprise. Its core architecture is built on a predefined schema (a "schema-on-write" approach), where data is cleaned, transformed, and structured before being loaded. This ensures data consistency and quality, making it ideal for company-wide analytics and business intelligence, which aligns perfectly with the scenario's requirements.

A. Data mart: This is a subset of a data warehouse, focused on a single business line or department, not the entire company.

C. Data silo: This refers to an isolated data repository, inaccessible to the rest of the organization, which is the opposite of an integrated, company-wide solution.

D. Data lake: A data lake stores raw data in its native format, including unstructured and semi-structured data, using a "schema-on-read" approach, which contradicts the requirement for a predefined data structure.

---

1. Microsoft Azure Documentation. (n.d.). What is a Data Warehouse?. Microsoft. Retrieved from https://azure.microsoft.com/en-us/resources/cloud-computing-dictionary/what-is-a-data-warehouse/.

Reference Detail: In the main definition

it states

"A data warehouse is a centralized repository of integrated data from one or more disparate sources. Data warehouses store current and historical data in one single place that are used for creating analytical reports for workers throughout the enterprise." This supports the "whole company" and "structured data" aspects.

2. Amazon Web Services (AWS) Documentation. (n.d.). What is a Data Lake?. AWS. Retrieved from https://aws.amazon.com/big-data/datalakes-and-analytics/what-is-a-data-lake/.

Reference Detail: The section "Data Lake vs. Data Warehouse" explicitly contrasts the two: "A data warehouse is a database optimized to analyze relational data... The data structure and schema are defined in advance... A data lake is different

because it stores relational data... and non-relational data... without the need to have a predefined schema." This clarifies why a data lake is incorrect.

3. Sawadogo

P.

& Darmont

J. (2021). Data Lakes: A Survey. The VLDB Journal

30(1)

137-166. https://doi.org/10.1007/s00778-020-00633-5

Reference Detail: Section 2.1

"Data Warehouses

" states: "Data warehouses (DWs) are mature decision-support systems that provide high-quality

cleansed

and consolidated data... DWs are based on a schema-on-write approach

where data must conform to a predefined schema before being loaded." This peer-reviewed source confirms the "predefined data structure" requirement for data warehouses.

4. Chaudhuri

S.

& Dayal

U. (1997). An overview of data warehousing and OLAP technology. ACM SIGMOD Record

26(1)

65-74. https://doi.org/10.1145/248603.248616

Reference Detail: Section 2

"What is a Data Warehouse?"

defines it as a "collection of decision support technologies

aimed at enabling the knowledge worker... to make better and faster decisions." It emphasizes the integration of data from multiple

heterogeneous sources into a single

structured repository.

The task is to "flag" customers based on a specific criterion: whether the Month is 'January'. This requires an operation that evaluates a condition for each row and returns a result based on whether that condition is true or false. Logical functions, such as IF or CASE, are designed specifically for this purpose. They test a condition (e.g., Month = 'January') and return a corresponding value (e.g., 1 for true, 0 for false), effectively creating a flag.

A. Statistical: Statistical functions are used to perform aggregations on a set of data, such as calculating SUM, AVERAGE, or COUNT, not for evaluating a row-by-row condition.

C. Mathematical: Mathematical functions perform arithmetic calculations like addition or multiplication on numeric data. They are not used for evaluating conditional statements on text values.

D. Date: Date functions are used to manipulate or extract components from date/time data types. The operation here is a logical comparison of a text string, not a date manipulation.

---

1. Microsoft Corporation. (n.d.). Logical functions in Data Analysis Expressions (DAX). Microsoft Learn. Retrieved from the official Microsoft documentation. The documentation for the IF function states

"The IF function checks a condition... and returns one value if the condition is TRUE

otherwise it returns a second value." This directly describes the process of flagging based on a condition.

2. The PostgreSQL Global Development Group. (2023). Chapter 9. Functions and Operators

Section 9.18. Conditional Expressions. PostgreSQL 16 Documentation. This official documentation describes the CASE expression

which "provide[s] a general conditional expression

similar to if/else statements." A common use is CASE WHEN month = 'January' THEN 'Yes' ELSE 'No' END

which is a direct implementation of the flagging logic required in the question.

3. VanderPlas

J. (2016). Python Data Science Handbook. O'Reilly Media. Chapter 3.02

"Data Indexing and Selection". In the context of data analysis with Python's Pandas library

this operation is achieved through boolean masking (e.g.

df['Month'] == 'January'). The handbook describes this as a logical operation that produces a boolean series

which can be used to flag

select

or filter data. This is a foundational concept taught in data science curricula.

Containerization is a method of operating-system-level virtualization used to deploy and run applications in isolated spaces called containers. A container packages an application's code along with all its necessary dependencies, such as libraries, system tools, and runtime environments. This creates a single, lightweight, executable software package that is portable and can run consistently and reliably across various computing environments, from a local machine to a cloud infrastructure.

A. Package manager: A package manager installs and manages software packages and their dependencies for a specific operating system or programming language, not for cross-environment portability.

B. Code repository: A code repository is a storage location for managing and tracking changes to source code (e.g., Git); it does not package the application for execution.

C. Virtual machine: A virtual machine emulates an entire computer system, including a full guest operating system, making it much more resource-intensive than just packaging an application and its libraries.

1. Docker Inc. (Official Vendor Documentation). "What is a Container?". Docker Resources. Accessed October 2023. The documentation states

"A container is a standard unit of software that packages up code and all its dependencies so the application runs quickly and reliably from one computing environment to another."

2. Bernstein

D. (2014). "Containers and Cloud: From LXC to Docker to Kubernetes." IEEE Cloud Computing

1(3)

pp. 81-84. DOI: 10.1109/MCC.2014.51. This paper (Section: "What Is a Container?") describes containers as a technology that packages an application with its runtime dependencies

enabling it to run on any infrastructure.

3. University of California

Berkeley. "CS 162: Operating Systems and Systems Programming

" Lecture 22: Virtual Machines. The course materials often contrast VMs with containers

defining containers as a lightweight alternative that bundles the application and its dependencies to run on a shared host OS kernel

ensuring environmental consistency.

This SQL query correctly addresses the user's request. The SELECT statement retrieves the cellular number, the carrier, and the sum of call durations. The SUM(callduration) is an aggregate function that calculates the total duration. When using an aggregate function, all other columns in the SELECT list that are not part of an aggregate function must be included in the GROUP BY clause. This query correctly groups the results by both cellularnumber and callednumbercarrier. Finally, the ORDER BY callednumbercarrier clause sorts the output as required.

A. The GROUP BY clause is incomplete. It must include all non-aggregated columns from the SELECT list, but it omits callednumbercarrier.

B. The GROUP BY clause incorrectly groups by callduration instead of the required cellularnumber, and it attempts to sort by a column not in the final output.

D. The GROUP BY clause is incorrect. It groups by callduration and omits the non-aggregated columns cellularnumber and callednumbercarrier from the SELECT list.

1. PostgreSQL 16 Documentation

Section 7.2.4. Grouping and Window Functions: "When GROUP BY is present

or if any aggregate functions are present

it is not valid for the SELECT list expressions to refer to ungrouped columns except within aggregate functions or when the ungrouped column is functionally dependent on the grouped columns

since there would otherwise be more than one possible value to return for an ungrouped column." This rule directly invalidates options A

B

and D

while confirming C is correct.

2. Stanford University

CS 145

Introduction to Databases

Lecture Notes on SQL Aggregation: The notes state the fundamental rule of SQL aggregation: "Attributes in SELECT list that are not aggregated must be in GROUP BY list." This principle is the basis for why option C is the only valid query among the choices.

3. MIT OpenCourseWare

6.830/6.814: Database Systems

Fall 2018

Lecture 2 Notes

Section 3.2 Aggregation: "The GROUP BY clause partitions the set of tuples into groups. The aggregate is then applied to each group... Any attribute that is not aggregated in the SELECT clause must appear in the GROUP BY clause." This academic source reinforces the standard SQL syntax and logic used in the correct answer.

The core issue is data staleness; the report is not reflecting recent updates from its source. The most direct and common cause for this is a failure in the data refresh process. This process is fundamentally dependent on the report's connection to the source database. The analyst should first verify the status of the database connection, including its configuration, credentials, and the success of the last scheduled data refresh. A failed, misconfigured, or unscheduled refresh is the most probable reason the new data is not appearing in the report.

A. Event log: An event log is a secondary diagnostic tool. It is more efficient to first check the component directly responsible for data updates (the connection/refresh) before searching through logs.

B. User privileges: The user can view the report, which means access privileges are not the problem. The issue is with the timeliness of the data itself, not the user's ability to access it.

D. Report corruption: Corruption would typically manifest as loading failures, visual errors, or an inability to open the report, not as the display of perfectly structured but outdated information.

1. Microsoft Power BI Documentation: In troubleshooting data refresh scenarios

the primary steps involve checking the data source connection. "A common reason for refresh failures is that Power BI can't sign in to your data sources... Check your data source credentials in the dataset settings." This confirms that the connection is the first point of inspection for stale data.

Source: Microsoft Learn

"Troubleshoot refresh scenarios

" Section: "Refresh fails because of sign-in problems."

2. Lemahieu

W.

vanden Broucke

S.

& Baesens

B. (2018). Principles of Data Management: Facilitating Information Sharing. Cambridge University Press. The text describes the Business Intelligence (BI) architecture where reports query a data warehouse or database. It states

"The data loading step is typically executed on a periodic basis (e.g.

nightly) to provide the business users with recent information." A failure in this loading or refresh process

which depends on the connection

is a primary cause of outdated reports.

Reference: Chapter 10

"Business Intelligence

" Section 10.2.1

"Data Warehouse Architecture."

3. Kimball

R.

& Ross

M. (2013). The Data Warehouse Toolkit: The Definitive Guide to Dimensional Modeling (3rd ed.). Wiley. The book emphasizes the importance of the Extract

Transform

and Load (ETL) system for populating the data warehouse that BI tools query. A failure in any part of this pipeline

starting with the connection to the source (Extract)

will result in stale data for end-users. Verifying the data pipeline's operational status is a primary troubleshooting step.

Reference: Chapter 2

"The Data Warehouse/Business Intelligence Architecture

" Section on "The ETL System."

Foundational models, particularly those built on the transformer architecture, represent a state-of-the-art approach for time-series forecasting. These large models are pre-trained on vast amounts of sequential data, enabling them to learn complex patterns, seasonality, and long-range dependencies. They can then be fine-tuned for specific forecasting tasks, often achieving superior accuracy compared to traditional statistical methods or other neural network types. Their ability to handle long input sequences makes them exceptionally well-suited for complex time-series problems.

A. Generative AI: This category's primary purpose is to create new content (e.g., text, images), not to perform predictive analysis on historical numerical data for forecasting.

C. Natural language processing: This field of AI is specifically concerned with understanding and processing human language, not the numerical, time-ordered data characteristic of time-series.

D. Robotic process automation: This technology automates rule-based business processes and digital tasks; it lacks the analytical and predictive capabilities required for forecasting.

---

1. Gruver

N.

et al. (2024). "Large Language Models are Zero-Shot Time Series Forecasters." Advances in Neural Information Processing Systems 36 (NeurIPS 2023). Section 1

Paragraph 2 discusses the limitations of existing methods and the potential for large

pre-trained models (i.e.

foundational models) to be adapted for time-series forecasting tasks.

2. Zhou

T.

et al. (2023). "One Fits All: A Generative Foundation Model for Time Series Analysis." Proceedings of the 40th International Conference on Machine Learning

PMLR 202

pp. 41933-41951. The abstract and introduction explicitly propose a "foundation model for time series analysis" to handle diverse datasets and tasks

including forecasting.

3. Rasul

K.

et al. (2024). "Chronos: Learning the Language of Time Series." Amazon Science. This official publication introduces Chronos

a family of pre-trained foundational models specifically designed for time-series forecasting

demonstrating a direct industry application of this concept. Available at: https://www.amazon.science/publications/chronos-learning-the-language-of-time-series

4. Vaswani

A.

et al. (2017). "Attention is All You Need." Advances in Neural Information Processing Systems 30 (NIPS 2017). This paper introduced the Transformer architecture

which is the core technology behind most modern foundational models and has been successfully adapted from its original NLP context to time-series forecasting.

A user information database contains Personally Identifiable Information (PII), which is data that can be used to identify a specific individual. Due to its potential for misuse and legal protection requirements (e.g., GDPR, CCPA), PII is classified as both Sensitive and Private. Sensitive data requires a high level of protection against unauthorized disclosure, and Private data is information about individuals that is not public. Implementing security triggers is a technical control used to audit access and prevent unauthorized actions on such high-risk data, which is consistent with these classifications.

A. Public: Public data has no confidentiality requirements and would not necessitate special security triggers for protection.

B. Open: Open data is intended for unrestricted access and sharing, which is the opposite of protected user information.

D. Non-Sensitive: This classification is for data with low impact if disclosed and would not warrant the implementation of security triggers.

F. Encrypted: Encryption is a security control or a data state, not a data classification level. Data is classified first, then controls like encryption are applied.

1. National Institute of Standards and Technology (NIST). (2010). Guide to Protecting the Confidentiality of Personally Identifiable Information (PII) (Special Publication 800-122). Section 2.2

"PII Confidentiality Impact Levels

" discusses how PII is categorized based on sensitivity and the potential impact of a breach

directly linking user information to the "Sensitive" concept.

2. Carnegie Mellon University

Information Security Office. Guidelines for Data Classification. This university guideline defines "Private" data as including names

email addresses

and other PII. It also defines "Sensitive" data as a category requiring a high level of security

which includes private data. This establishes the use of both terms for user information.

3. Oracle Corporation. Oracle Database Security Guide

21c. Chapter 1

"Introduction to Oracle Database Security

" discusses the importance of identifying sensitive data within the database to apply appropriate security controls. This aligns with the practice of classifying user information as sensitive to justify implementing measures like security triggers.

User Acceptance Testing (UAT) is the final phase in the software testing process. It is performed by the end-users or client representatives to validate that the software system meets the agreed-upon business requirements and is fit for its intended purpose from a user's perspective. The primary goal of UAT is to gain confidence and formal acceptance from the user that the system is ready for operational use. This involves testing the software in a real-world or simulated operational environment to ensure it can handle the required tasks and workflows.

A. UAT is a final validation stage that occurs before deployment, not at the beginning of the development process.

B. Ensuring different parts of an application work together is the purpose of Integration Testing, not UAT.

C. Reviewing crashes and deploying patches is part of the post-deployment maintenance and support phase, which happens after UAT.

1. ISO/IEC/IEEE 29119-1:2013

Software and systems engineering — Software testing — Part 1: Concepts and definitions. Section 4.2

"Test levels

" defines acceptance testing as a test level focused on determining whether to accept the system

often performed by customers or users based on their requirements.

2. Bourque

P.

& Fairley

R. E. (Eds.). (2014). Guide to the Software Engineering Body of Knowledge (SWEBOK)

Version 3.0. IEEE Computer Society. In Chapter 4

"Software Testing

" Section 2.1.3

"Acceptance Testing

" it is stated that this testing determines if a system satisfies its acceptance criteria

enabling the customer to determine whether to accept the system.

3. University of Toronto

Department of Computer Science. (2021). CSC301: Introduction to Software Engineering - Lecture 10: Software Testing. The course material describes User Acceptance Testing (UAT) as the process where clients/users test to confirm the system meets their needs

serving as a final verification of required business functions.

In standard SQL, NULL is a special marker indicating that a data value does not exist. It is not equivalent to zero or an empty string. Standard comparison operators like = cannot be used to test for NULL because the result of any arithmetic comparison with NULL is also NULL (or unknown), which evaluates to false in a WHERE clause. The correct, standard-compliant operator to specifically test for the absence of a value is IS NULL. This syntax correctly identifies and filters for all records where the specified column is empty.

A. WHERE [ColumnName] = NULL is incorrect because the = operator cannot be used to test for NULL values; this expression will not return any rows.

C. WHERE [ColumnName] IS NOT NULL is the opposite of the required logic; it returns records where the field has a value, not where it is empty.

D. WHERE [ColumnName] = 'NULL' searches for the literal text string 'NULL', not for fields that are actually empty (i.e., have the NULL state).

1. PostgreSQL 16 Documentation. (2023). 9.2. Comparison Functions and Operators. In the section detailing comparison operators

it states

"Do not write expression = NULL because NULL is not 'equal to' NULL... To test for nullness

use the constructs IS NULL and IS NOT NULL." Retrieved from postgresql.org/docs/16/functions-comparison.html.

2. Microsoft Transact-SQL (T-SQL) Documentation. (2023). IS NULL (Transact-SQL). Microsoft Learn. This official documentation specifies

"To determine whether an expression is NULL

use IS NULL or IS NOT NULL instead of comparison operators (such as = or !=)." Retrieved from learn.microsoft.com/en-us/sql/t-sql/queries/is-null-transact-sql.

3. Ullman

J. D.

& Widom

J. (n.d.). CS145 Introduction to Databases

Lecture 4: SQL 1. Stanford University. In the lecture slides covering SQL basics

slide 28 ("A Subtlety About NULLs") explicitly states: "Test for NULL using IS NULL and IS NOT NULL

not = or ." Retrieved from web.stanford.edu/class/cs145/slides/04-sql1.pdf.

The primary requirement is to visualize the "composition of total sales amounts" across different channels. A pie chart is the most suitable visualization for this purpose as it is specifically designed to show part-to-whole relationships. Each slice of the pie would represent a sales channel (Digital, Store, Online), and the size of the slice would be proportional to its contribution to the total sales, clearly illustrating the composition.

A. Pivot table: A pivot table is a tool for data summarization in a tabular format, not a graphical visualization designed to show proportional composition.

C. KPI card: A Key Performance Indicator (KPI) card is used to display a single, aggregate metric (e.g., total sales), not the breakdown or composition of that metric across categories.

D. Box plot: A box plot is used to display the statistical distribution of numerical data (e.g., median, quartiles, outliers), not the proportional contribution of categories to a total.

---

1. Microsoft Power BI Documentation. (n.d.). Pie charts in Power BI. Microsoft Docs. Retrieved from the official Power BI documentation on visualization types. In the section "When to use a pie chart

" it states

"Pie charts are a great choice... To show proportions." This directly aligns with the question's need to show the composition of total sales.

2. Wilke

C. O. (2019). Fundamentals of Data Visualization: A Primer on Making Informative and Compelling Figures. O'Reilly Media. In Chapter 10

"Visualizing proportions

" Section 10.1

"Pie charts

" the author states

"A pie chart represents the whole of a given quantity as a circle

and the parts as wedges of that circle. The parts are sized by their relative contribution to the whole." This is a standard use case taught in data visualization curricula at institutions like the University of Texas at Austin

where the author is a professor.

3. University of Washington. (n.d.). Data Visualization: Basic Principles. Course materials for DATA 512: Data Visualization. The course materials explain that pie charts are used to show the "relative shares of categories that sum to a whole

" which is precisely the task of showing the composition of sales channels.

The most efficient and standard method for this task is to work with structured data formats that Python's data analysis libraries, like pandas, are optimized for. Web APIs commonly return data in JSON (JavaScript Object Notation) format. This structured format can be directly and easily parsed into a pandas DataFrame, preserving data types and relationships. Similarly, data queried from a database can be loaded directly into a DataFrame. Once both data sources are represented as DataFrames, pandas provides powerful and optimized merge or join functions to combine them based on common keys. This approach minimizes intermediate conversion steps and avoids the loss of data type information.

A. varchar is a database-specific data type, not a universal data interchange format. Converting all data to this string-like type would unnecessarily discard valuable numeric and date-type information.

C. Converting structured API and database data to a flat TXT file is an inefficient intermediate step that adds complexity and risks data integrity issues during parsing.

D. Converting all data to a generic string format is a poor practice as it loses essential data type information (e.g., integers, floats), making subsequent numerical analysis and joins difficult.

1. Pandas Official Documentation

IO Tools (JSON): The documentation explicitly details the pandas.readjson function for converting a JSON string directly into a pandas DataFrame. This is the standard method for handling API data.

Source: Pandas Development Team. (2023). pandas-docs: IO Tools (Text

CSV

HDF5

…). Section: JSON. Retrieved from https://pandas.pydata.org/pandas-docs/stable/userguide/io.html#json

2. Pandas Official Documentation

IO Tools (SQL): The documentation shows the pandas.readsql function

which is designed to read the result of a SQL query directly into a DataFrame

preserving data types from the database where possible.

Source: Pandas Development Team. (2023). pandas-docs: IO Tools (Text

CSV

HDF5

…). Section: SQL queries. Retrieved from https://pandas.pydata.org/pandas-docs/stable/userguide/io.html#sql-queries

3. University of Illinois Urbana-Champaign

Courseware: In the "Data Visualization" course materials

the standard workflow for handling web data is demonstrated. It involves making an API request

parsing the resulting JSON response

and loading it into a data structure like a pandas DataFrame for analysis.

Source: University of Illinois Urbana-Champaign. (2022). CS 307: Working with APIs and JSON. Week 3 Notes. Retrieved from https://uiuc-ischool-dataviz.github.io/spring2022/week03/workingwithapisandjson.html (Specifically

the section on parsing JSON and creating DataFrames).

The analyst has performed initial, internal troubleshooting steps: using print statements for debugging, conducting a peer review, and verifying software compatibility and versioning. When these methods fail, the next logical step in a standard troubleshooting methodology is to broaden the search for a solution by consulting external resources. Researching the issue online provides access to official documentation, knowledge bases, and community forums where similar problems and their resolutions are often documented. This step is a crucial part of due diligence before escalating the issue or resorting to workarounds.

A. This is a temporary workaround, not a troubleshooting step to solve the current problem. It avoids the issue rather than addressing it.

B. It is premature to declare a software bug and contact IT. The analyst should first exhaust self-service resources, as the issue might be a known problem with a documented solution.

C. Escalating to a manager is inappropriate for a technical problem at this stage. This step should only be taken after standard technical troubleshooting and escalation paths have been exhausted.

1. Sommerville

I. (2016). Software Engineering (10th ed.). Pearson.

In Chapter 8

Section 8.4 on "Debugging

" the text describes the process of locating and repairing defects. A key part of locating a defect is information gathering. When initial hypotheses fail

a developer must gather more information

which in a modern context includes searching for known issues in external knowledge bases before escalating. This aligns with researching the problem online.

2. Zeller

A. (2009). Why Programs Fail: A Guide to Systematic Debugging (2nd ed.). Morgan Kaufmann.

Chapter 1

"Isolating Failure Causes

" emphasizes a scientific method for debugging: observe

hypothesize

test

and repeat. When a programmer's own hypotheses are exhausted (as in the scenario)

the next step is to gather more data or see if others have already solved the problem. The book implicitly supports consulting external documentation and community knowledge as part of the observation/information-gathering phase.

3. Ko

A. J.

& Myers

B. A. (2004). A framework and methodology for studying the causes of software errors in programming systems. Journal of Visual Languages & Computing

15(1)

41-84. https://doi.org/10.1016/j.jvlc.2003.07.004

This paper discusses the cognitive processes of debugging. It highlights that a major part of resolving errors involves "information seeking." After failing to find information through code inspection (as the analyst did)

a programmer naturally progresses to external information sources like documentation or online communities. This supports online research as the next logical action.

The configuration aligns data visualization best practices with the dashboard's requirements.

• The "Power Outages Enterprise-wide" title matches the director's need to monitor the "entire state."
• For "County Outages," a map is the most effective visualization for displaying data distributed across a geographic area.
• For "Status," the grouped bar chart is ideal as it allows for the comparison of categorical data (Closed, In Progress, Reported) across multiple groups (Counties).
• For "Number of Outages for the Quarter," the pie chart is used to represent a part-to-whole relationship, showing how the total quarterly outages are distributed among the counties.

Microsoft. (n.d.). Basic concepts for designers in the Power BI service. Microsoft Learn. Retrieved November 7, 2025.

Reference: In the section "Visualizations in Power BI reports," Microsoft documentation guides the use of "Maps" for spatial data, "Bar charts" for comparing values across categories, and "Pie charts" for showing the relationship of parts to a whole. This supports the selection of the map for counties, the bar chart for status, and the pie chart for the quarterly breakdown.

University of Washington. (n.d.). Data Visualization: A Guide to Good Practice. UW Libraries.

Reference: This guide discusses chart selection, noting that "Chloropleth maps are used to show data aggregated over geographical regions" (supporting the choice for "County Outages") and "Bar charts are used to compare values across categories" (supporting the choice for "Status").

IEEE. (2011). Information Visualization. In VIS '11: Proceedings of the IEEE VIS 2011.

Reference: Academic principles of information visualization emphasize selecting a visual encoding that matches the user's task. The task "identify where... resources should be allocated" (geographic) is directly supported by a map. The task of monitoring "status" (comparison) is supported by a bar chart.

The SQL UNION operator is designed to combine the result sets of two or more SELECT statements into a single result set. A key characteristic of the UNION operator is that it automatically removes duplicate rows from the final result. The analyst's task is to merge records from the "Employees" and "Managers" tables and eliminate any duplicates between them. The statement SELECT FROM Employees UNION SELECT FROM Managers correctly performs this operation by vertically appending the two tables and filtering out any identical rows, thus achieving the desired outcome.

A. UNION ALL combines result sets but includes all rows, including duplicates, which fails to meet the requirement to remove them.

C. The JOIN syntax is invalid, and its purpose is to combine columns from different tables based on a related key, not to combine rows.

D. CROSS JOIN creates a Cartesian product, pairing every row from the first table with every row from the second, which is not the required operation.

1. PostgreSQL 16 Documentation. (n.d.). 7.4. Combining Queries. PostgreSQL. Retrieved from https://www.postgresql.org/docs/16/queries-union.html. In Section 7.4

the documentation states

"The UNION operator computes the set union of the rows returned by the involved SELECT statements. A row is in the set union of two result sets if it appears in at least one of the result sets. ... Duplicate rows are eliminated from the result..."

2. Oracle Database SQL Language Reference

23c. (n.d.). Set Operators. Oracle. Retrieved from https://docs.oracle.com/en/database/oracle/oracle-database/23/sqlrf/Set-Operators.html#GUID-B0F24731-B6B5-465F-940E-25D617553A0E. The documentation specifies

"The UNION operator returns the distinct rows from the result sets of the two queries." It contrasts this with UNION ALL

which "does not eliminate duplicate rows."

3. Ullman

J. D.

& Widom

J. (n.d.). CS145 Introduction to Databases

Lecture Notes: SQL 2. Stanford University. Retrieved from https://web.stanford.edu/class/cs145/notes/SQL-2.pdf. On page 2

under "Union

Intersection

Difference

" the notes explain that (SELECT ... FROM ... WHERE ...) UNION (SELECT ... FROM ... WHERE ...) computes the set union and that "UNION removes duplicates."

A scatter plot is the most suitable visualization for examining the relationship or correlation between two continuous numerical variables. In this scenario, 'age' and 'income' are both continuous variables. Each employee can be represented as a point on the graph, with their age plotted on the x-axis and their income on the y-axis. This visual representation allows the human resources department to easily identify patterns, trends (e.g., does income increase with age?), and outliers within the dataset.

B. Area plot: An area plot is typically used to show how a quantitative value changes over a continuous dimension, such as time, and is not suitable for correlating two distinct variables.

C. Bar chart: A bar chart is designed to compare values across discrete categories, not to illustrate the relationship between two continuous variables like age and income.

D. Pie chart: A pie chart is used to display the proportional composition of a single categorical variable, representing parts of a whole, and cannot show a relationship between two variables.

1. Wilke

C. O. (2019). Fundamentals of Data Visualization. O'Reilly Media

Inc. In Chapter 12

"Visualizing associations among two or more quantitative variables

" the text states

"The standard visualization for an association between two quantitative variables is the scatterplot." The chapter proceeds to provide examples using this chart type.

2. University of Virginia Library

Research Data Services. (n.d.). Data Visualization: The Scatterplot. Retrieved from the University of Virginia Library website. The guide specifies

"A scatterplot is a type of data visualization that shows the relationship between two numerical variables... Scatterplots are used to observe and show relationships between two numeric variables."

3. Healy

K. (2018). Data Visualization: A Practical Introduction. Princeton University Press. In Chapter 3

"Make a Plot

" the section on scatterplots explains

"Scatterplots are the default choice for visualizing the relationship between two continuous variables" (p. 56).

The dataset contains both "dates of birth" and "employees’ birth months." The birth month can be directly derived or calculated from the date of birth. Storing data that can be derived from other fields within the same record is a form of data redundancy. This practice is inefficient as it increases storage needs and, more importantly, creates a risk of inconsistency. For example, if the date of birth is corrected, the birth month field might not be updated, leading to conflicting information within the same record. Identifying and resolving such redundancy is a fundamental step in data cleaning and governance.

B. Equivalence: This refers to different data values representing the same real-world entity (e.g., 'St.' and 'Street'), which is not the issue here.

C. Parallel: This term is not a standard classification for data inconsistency; it typically relates to system architecture or data processing methods.

D. Duplication: This usually refers to entire records (rows) being repeated in a dataset, not redundant information within a single record.

---

1. Elmasri

R.

& Navathe

S. B. (2017). Fundamentals of Database Systems (7th ed.). Pearson. In Chapter 15

"Basics of Functional Dependencies and Normalization for Relational Databases

" the authors discuss how storing derivable values (e.g.

a person's age from their birth date) introduces redundancy and can lead to update anomalies (Section 15.2.1). The presence of both 'date of birth' and 'birth month' is a direct example of this principle.

2. DAMA International. (2017). DAMA-DMBOK: Data Management Body of Knowledge (2nd ed.). Technics Publications. Chapter 5

"Data Modeling and Design

" emphasizes that a key goal of sound data modeling is to reduce data redundancy. It defines redundancy as the unnecessary repetition of data

which can be within a single record (as a derivable attribute) or across a database.

3. Stanford University. (2016). CS145: Introduction to Databases - Lecture 8: Functional Dependencies and Normalization. Retrieved from web.stanford.edu/class/cs145/slides/08-fds-and-normalization.pdf. Slide 30 discusses redundancy as information being "repeated unnecessarily" and illustrates the update

insertion

and deletion anomalies that arise from it

which is the core problem with having both a date of birth and a derivable birth month.

The UNION operator in SQL is used to combine the result sets of two or more SELECT statements into a single result set. It appends the rows from one table to another. For this operation to work, the tables must have the same number of columns, and the corresponding columns must have compatible data types. In this scenario, both the Managers and Non-managers tables share an identical structure, making UNION the correct method to create a single, consolidated report of all employees by stacking the rows from both tables vertically.

A. Group: This is used with aggregate functions (e.g., COUNT, SUM) to summarize data into groups, not to combine tables.

B. Join: This combines tables horizontally based on a related column, merging columns from multiple tables into a single row.

D. Nested: This refers to a subquery (a query within another query) and describes a query's structure, not the specific operation for combining these tables.

1. Microsoft SQL Server Documentation. (2023). UNION (Transact-SQL). Microsoft Docs. In the "Arguments" section

it is stated that UNION "Specifies that multiple result sets are to be combined and returned as a single result set." The basic example shows combining two tables with identical column structures.

2. PostgreSQL 16 Documentation. (2023). 7.4. Combining Queries. In this section

the documentation states

"The UNION operator computes the set union of the rows returned by the involved SELECT statements... The two SELECT statements that represent the direct operands of the UNION must produce the same number of columns

and corresponding columns must be of compatible data types."

3. Ullman

J. D.

& Widom

J. (n.d.). CS145 Introduction to Databases Course Notes: More SQL. Stanford University. In the section on "Union

Intersection

and Difference of Relations

" the notes specify that UNION is the standard set-union operator for SQL

requiring the tables to be "union-compatible

" which the tables in the question are.

The most effective way to find all customers who have not placed an order is to use a LEFT JOIN. This query starts with the Customers table (the "left" table) and joins it to the Products table. A LEFT JOIN returns all records from the Customers table, regardless of whether there is a matching record in the Products table. If a customer has no orders, the columns from the Products table for that customer's row will be populated with NULL. The WHERE P.CustomerID IS NULL clause then filters the results to show only those rows where no match was found, effectively isolating the customers without orders.

B. This query uses an INNER JOIN, which would only return customers who have placed an order. It also contains multiple syntax errors, such as an invalid join condition and the use of an aggregate function (COUNT) in a WHERE clause.

C. This query also incorrectly uses an INNER JOIN, which would filter out customers without orders from the start. The WHERE clause contains a syntactically invalid subquery that attempts to use a boolean expression.

D. This query performs a LEFT JOIN from Products to Customers and filters for non-null CustomerIDs. This would return all orders and their associated customer information, which is the opposite of the desired outcome.

---

1. Silberschatz

A.

Korth

H. F.

& Sudarshan

S. (2020). Database System Concepts (7th ed.). McGraw-Hill. In Chapter 4

"Intermediate SQL

" the section on "Outer-Join Expressions" describes how a LEFT OUTER JOIN preserves tuples from the left relation that do not have matching tuples in the right relation

extending them with null values. This is the foundational principle for the correct answer's logic.

2. PostgreSQL 16 Documentation. (2023). "7.2.2. JOIN Clauses". The documentation for LEFT OUTER JOIN states: "For each row R1 of T1

the joined table has a row for each row in T2 that satisfies the join condition with R1. ... If no join partner exists in T2

a row is generated with null values in the T2 columns." This directly supports the mechanism of using WHERE ... IS NULL to find non-matching rows.

3. SQLite Documentation. "The LEFT and CROSS JOIN keywords". The official documentation explains: "If the ON or USING clause of a LEFT JOIN is not satisfied for some row in the left-hand table

then a single row is output for that left-hand table row with NULL values in the columns that would have been filled from the right-hand table." This confirms the behavior exploited by the correct query.

A data lake is a centralized storage repository designed to hold vast amounts of raw data in its native format. The fundamental principle of a data lake is to ingest and store data "as-is" from various sources without prior transformation or structuring (a concept known as schema-on-read). This approach preserves the original, unaltered data, making it available for diverse and future analytical purposes. In contrast, other repositories like data warehouses require data to be cleaned, transformed, and structured into a predefined schema before it is loaded (schema-on-write), thereby altering it from its original state.

B. Data warehouse: Stores structured and transformed data that has been processed (ETL/ELT) to fit a predefined schema, meaning it is not in its original, unaltered form.

C. Data table: A specific structure within a database that holds organized, structured data, which by definition has been processed to fit the table's schema.

D. Data factory: A cloud service for orchestrating and automating data movement and transformation (ETL/ELT pipelines); it is a processing tool, not a storage repository.

1. Microsoft Azure Documentation. (2023). What is Azure Data Lake?. "A data lake is a centralized repository that allows you to store all your structured and unstructured data at any scale. You can store your data as-is

without having to first structure the data..."

2. Amazon Web Services (AWS) Documentation. (n.d.). What is a Data Lake?. "A data lake is a centralized repository that allows you to store all your structured and unstructured data at any scale. You can store your data as is

without having to first structure the data..."

3. Nargesian

F.

Ghafori

E.

Toosi

A.

& van der Aalst

W. M. P. (2019). Data Lake Governance: A Survey. ACM Computing Surveys

52(4)

Article 80

p. 2. "A data lake is a storage repository that holds a vast amount of raw data in its native format until it is needed." DOI: https://doi.org/10.1145/3342260

4. Hai

R.

Geisler

S.

& Quix

C. (2016). Constance: An Intelligent Data Lake System. In Proceedings of the 2016 International Conference on Management of Data (SIGMOD '16). Association for Computing Machinery

New York

NY

USA

2097–2100. Section 2

"Data Lakes". "Data lakes keep the data in its raw format and only apply a schema when the data is read (schema-on-read)." DOI: https://doi.org/10.1145/2882903.2903740

Data imported from a flat file, such as a CSV or TXT file, is often interpreted as a generic text or string data type by default. To perform date-specific calculations, like finding the duration between two dates, the string data must first be explicitly converted (parsed) into a date or datetime data type. Once the data is in the correct format, built-in date functions can be leveraged to accurately calculate the difference in days. This approach is standard practice as it correctly handles complexities like leap years and different date formats, ensuring reliable and accurate results.

B. Validation with logical functions only checks if the data conforms to a pattern; it does not change the data type, which is necessary for date arithmetic.

C. Applying date functions directly to unconverted string data will almost always result in a data type mismatch error or produce incorrect results.

D. Transforming dates into a single numerical value for mathematical operations is unnecessarily complex, error-prone, and bypasses optimized, built-in date functions.

1. PostgreSQL 14 Documentation

Chapter 9. Data Types

Section 9.9. Date/Time Functions and Operators. The documentation specifies the use of functions like TODATE(text

text) to convert strings to a date type. Following this

it details date arithmetic

stating

"Subtraction of a date from a date results in an integer (the number of days)." This two-step process (convert

then operate) aligns directly with the correct answer.

2. Pandas 2.2.2 Documentation

API Reference

pandas.todatetime. The official documentation for this core function in the Python pandas library states its purpose is to "Convert argument to datetime." This is a prerequisite for performing time-series operations

such as subtracting dates to get a Timedelta object

from which the number of days can be extracted.

3. Wickham

H.

& Grolemund

G. (2017). R for Data Science. O'Reilly Media. Chapter 16

"Dates and times with lubridate." This peer-reviewed book

widely used in university curricula

explains

"The first step is to parse the date-time so that you have a numeric representation of the time. [...] Once you have a date-time object

you can start doing useful work with it..." This reinforces the principle that parsing/conversion is the essential first step.

4. University of California

Berkeley

Data 8: Foundations of Data Science Course Textbook

Chapter 6.2

"Manipulating Columns." The textbook demonstrates that to perform arithmetic operations

columns must be of the correct numerical type. It shows the explicit conversion of string data to numbers using int() or float(). The same fundamental principle applies to dates: they must be converted to a date type before date-specific functions or arithmetic can be used.

The scenario describes a need for custom reports for "future executive meetings." This phrasing indicates a predictable, repeated requirement that will occur on a regular schedule (e.g., weekly, monthly, or quarterly). A recurring delivery method automates the generation and distribution of these reports at predefined intervals. This approach is the most efficient and reliable way to ensure that the CEO and other executives consistently receive the necessary information ahead of each scheduled meeting, eliminating the need for manual creation each time.

A. Ad hoc: This method is for one-time, unscheduled, or spontaneous report requests. The requirement has evolved from a single instance to a predictable, ongoing need for future meetings.

B. Real-time: This delivery method provides continuous, up-to-the-second data streams. It is unnecessary for strategic executive meetings, which typically review performance over a specific period (e.g., the last quarter).

D. Self-service: This involves providing executives with tools to create their own reports. The scenario explicitly states the analyst is responsible for generating and delivering the custom reports.

1. Sharda

R.

Delen

D.

& Turban

E. (2020). Business Intelligence

Analytics

and Data Science: A Managerial Perspective (4th ed.). Pearson. In Chapter 3

"Descriptive Analytics II: Business Intelligence and Data Warehousing

" the text distinguishes between different types of reports. It describes recurring (or scheduled) reports as those produced at predefined intervals to support routine informational needs

which directly aligns with the scenario of preparing for regular executive meetings.

2. Kimball

R.

& Ross

M. (2013). The Data Warehouse Toolkit: The Definitive Guide to Dimensional Modeling (3rd ed.). Wiley. Chapter 1

"Data Warehousing

Business Intelligence

and Dimensional Modeling Primer

" discusses the delivery of information through standard

pre-defined reports. The need for a consistent report for future meetings fits the description of a standard report delivered on a recurring schedule.

3. University of Washington. (n.d.). Introduction to Business Intelligence. Foster School of Business Courseware for MSIS 526. The course materials differentiate reporting styles

defining recurring reports as automated reports generated on a schedule (daily

weekly

monthly) for consistent operational and strategic oversight

contrasting them with ad-hoc reports for specific

unplanned questions.

The term "on-premises" refers to an IT infrastructure environment where the computing resources, such as servers, storage, and networking hardware, are physically located within an organization's own facilities (e.g., its headquarters or a private data center). The organization is solely responsible for the maintenance, operation, and security of this infrastructure. The scenario described, where the company's entire server environment is at its headquarters, perfectly matches this definition.

A. Cloud: A cloud environment involves infrastructure hosted by a third-party provider and accessed over the internet, not located at the company's own headquarters.

C. Public: This refers to a public cloud model, a specific type of cloud computing where services are delivered over the public internet and shared among multiple organizations.

D. Hybrid: A hybrid environment is a mix of on-premises infrastructure and a public or private cloud. The question states the entire environment is at one location, ruling this out.

---

1. National Institute of Standards and Technology (NIST). (2011). The NIST Definition of Cloud Computing (Special Publication 800-145). U.S. Department of Commerce. In Section 2

"Deployment Models

" the document defines cloud models (Public

Private

Hybrid) which are distinct from the traditional model where an organization manages its own infrastructure on its own property

i.e.

on-premises. The concept of a private cloud "on-premises" is discussed

but the general term for the location itself is on-premises. (Available at: https://doi.org/10.6028/NIST.SP.800-145)

2. Microsoft Azure Documentation. (n.d.). What is on-premises? Microsoft. The documentation defines the term explicitly: "On-premises means the computing infrastructure is located at a company’s own business or facility... The company is responsible for all the hardware

software

and data on-premises." This directly supports the correct answer. (Available from Microsoft's official Azure documentation portal).

3. Armbrust

M.

Fox

A.

Griffith

R.

Joseph

A. D.

Katz

R.

Konwinski

A.

Lee

G.

Patterson

D.

Rabkin

A.

Stoica

I.

& Zaharia

M. (2009). Above the Clouds: A Berkeley View of Cloud Computing (Technical Report No. UCB/EECS-2009-28). University of California at Berkeley. Section 2.1 discusses the distinction between cloud computing and traditional in-house data centers (i.e.

on-premises)

highlighting that on-premises requires the organization to manage its own hardware and software infrastructure.

After confirming there are no system-wide technical or operational issues (e.g., server load, network latency, or connection failures), the most logical next step is to investigate the report's specific configuration. Slow performance in a customized report is frequently caused by the query processing an unnecessarily large volume of data. Applying or verifying appropriate filters is the most direct method to reduce the dataset size that the report must load and render, which is a primary technique for optimizing query performance and resolving slow load times.

B. Check data source connections: This would be considered a technical or operational issue, which the analyst has already ruled out as part of the initial system review.

C. Check for data structure changes in the report: While a schema change can impact performance, it more commonly results in a query error rather than just slowness. Data volume is a more frequent cause of slow loading.

D. Check whether other peers have the same issue: This is a scoping step to determine if an issue is isolated or systemic. Since the analyst has already ruled out systemic issues, this step is redundant.

1. Microsoft Power BI Documentation. (2023). Optimization guide for Power BI. In the section "Report layout

design

and interactions

" the guide emphasizes limiting the amount of data a report needs to display. It states

"The more data that a visual needs to display

the slower that visual is to load... The best way to avoid this is to summarize the data... You can also use slicers and filters to limit the data." This directly supports checking filters to resolve performance issues.

2. Tableau Help. (2024). Optimize Workbook Performance. In the section "Filter your data

" the documentation advises

"One of the easiest ways to improve workbook performance is to filter the data that you are working with... When you are working with a large amount of data

filters may be your best friend." This confirms that filtering is a primary method for addressing slow performance.

3. Kimball

R.

& Ross

M. (2013). The Data Warehouse Toolkit: The Definitive Guide to Dimensional Modeling (3rd ed.). Wiley. Chapter 2

"The Business Dimensional Lifecycle

" discusses the importance of query performance. The principles of dimensional modeling are designed to facilitate high-performance "slicing and dicing

" which is functionally achieved through filtering in reports to constrain queries and return smaller

manageable result sets (p. 40-41).

4. Chaudhuri

S.

& Dayal

U. (1997). An overview of data warehousing and OLAP technology. ACM Sigmod Record

26(1)

65-74. This foundational paper on data warehousing explains that a key OLAP operation is the "slice

" which is "a selection on one dimension of the given cube" (p. 68). This operation

equivalent to applying a filter

is fundamental to reducing the data volume for analysis and improving performance. DOI: https://doi.org/10.1145/248603.248616

The analyst's objective is to examine five years of historical data to "identify patterns that can be studied further." This task directly aligns with diagnostic analytics. Diagnostic analytics focuses on understanding the root cause of past events by answering the question, "Why did it happen?" It involves techniques like data discovery, drill-down, and correlation analysis to uncover patterns and dependencies in historical data, which is precisely what the analyst intends to do to understand employee retention trends.

A. Predictive: This method is used to forecast future events (e.g., "What is the likely retention rate next year?"), not to analyze the cause of past events.

B. Prescriptive: This method recommends specific actions to achieve a desired outcome (e.g., "What steps should we take to improve retention?"), which is a step beyond the current analysis phase.

D. Descriptive: This method summarizes past data to show what happened (e.g., "What was our retention rate last year?"). While a dashboard is descriptive, the goal of finding patterns for study is a deeper, diagnostic activity.

---

1. University Courseware:

Carnegie Mellon University

Heinz College of Information Systems and Public Policy. "95-865 Unstructured Data Analytics - Lecture 1: Introduction to Business Analytics." This courseware defines the four types of analytics. It specifies that diagnostic analytics is used to answer "Why did it happen?" by examining past data to find root causes and patterns

distinguishing it from descriptive analytics ("What happened?").

2. Peer-Reviewed Academic Publication:

Delen

D.

& Ram

S. (2018). Research challenges and opportunities in business analytics. Journal of Business Analytics

1(1)

2-12. https://doi.org/10.1080/2573234X.2018.1434229. On page 4

the authors define diagnostic analytics as the methods used "to understand the underlying causes of a specific observed phenomenon

" which directly corresponds to the analyst's task of identifying patterns for further study in retention data.

3. Official Vendor Documentation:

Microsoft. (2023). "Get started with Power BI." In the Power BI service documentation

the "Basic concepts for designers" section explains how to move from reports (descriptive) to insights. The process of using features like "Analyze" to "explain the increase/decrease" in charts is a practical application of diagnostic analytics

which involves finding patterns and reasons within the data.