Free CompTIA Data+ Sample Questions

Correct answer: D — Tableau

• A. Power BI is a strong drag-and-drop BI tool, but it is best known for deep Microsoft 365/Azure integration rather than the connector breadth and chart-gallery reputation described here.
• B. Looker centers on a governed semantic modeling layer (LookML) for consistent metrics rather than freeform drag-and-drop chart building, so it does not match this description.
• C. Anaconda is a Python/R distribution and package manager, not a business intelligence visualization tool.
• D. Tableau is widely recognized for its broad connector library and drag-and-drop chart gallery aimed at fast, ad hoc visual exploration, matching the scenario.

Correct answers: A, B, C — DBeaver; Toad; MySQL Workbench

• A. DBeaver is a generic GUI client for connecting to, browsing, and querying a wide range of relational and non-relational databases, making it a DBMS management tool.
• B. Toad is a commercial GUI for writing queries, tuning performance, and administering database schemas, making it a DBMS management tool.
• C. MySQL Workbench is Oracle's official GUI for designing, querying, and administering MySQL databases, making it a DBMS management tool.
• D. tidyverse is a collection of R packages for data wrangling and visualization, not a database management GUI.
• E. Looker is a business intelligence platform for building governed dashboards, not a database administration GUI.
• F. Scala is a general-purpose programming language, not a GUI tool for managing databases.

Correct answer: A — Generative AI

• A. Generative AI is correct because the tool is producing new, original text tailored to each email rather than following a fixed template. Drafting a full reply from scratch is exactly the kind of content-creation task this category of AI performs.
• B. This is incorrect because robotic process automation automates repetitive clicks and data transfers between systems; it does not compose original written language in response to varied input.
• C. This is incorrect because image classification categorizes visual inputs like photos, and the scenario involves generating written text, not analyzing images.
• D. This is incorrect because a fixed set of hand-coded rules cannot produce a unique, context-appropriate reply for every different email; that requires a model that generates language dynamically.

Correct answers: A, B, E — The tool can produce fluent, human-like text without a person writing each sentence from scratch; The tool can occasionally generate fabricated facts or citations that sound plausible but are not accurate; The tool's output quality depends heavily on how clearly and specifically the prompt describes the desired outline

• A. This is correct because generating fluent, original text without a person drafting every sentence is a core, well-documented capability of generative AI tools like the one being evaluated.
• B. This is correct because generative AI models are known to sometimes hallucinate, producing confident but fabricated facts or citations, which the marketing team should account for when reviewing drafts.
• C. This is incorrect because generative AI models generalize from broad pretraining and do not need a dedicated labeled example for every specific topic before producing text on it.
• D. This is incorrect because clicking through existing software screens without generating new content describes robotic process automation, not the generative AI tool being evaluated.
• E. This is correct because the specificity and clarity of the prompt strongly influences how relevant and useful the generated outline turns out to be, which is a practical consideration for the team.
• F. This is incorrect because no generative AI tool guarantees perfectly factual output, and claiming exclusive training on peer-reviewed sources does not eliminate the risk of fabricated or inaccurate content.

Correct answer: A — File storage

• A. Organizing data in nested folders and serving it to multiple users over a network file-sharing protocol like SMB or NFS is exactly what this storage type provides.
• B. A raw volume formatted for one server's file system does not natively support simultaneous folder browsing by many employees over a network protocol.
• C. Files addressed by a unique identifier over an API do not present the nested folder browsing experience the department wants from their office applications.
• D. Storage confined to a single device cannot be shared across the accounting department's employees or accessed over the network.

Correct answer: C — Containerization

• A. Each virtual machine runs its own full guest operating system, which is the overhead the team specifically wants to avoid.
• B. Combining on-premises and provider infrastructure addresses where workloads run, not how an application and its dependencies are packaged for consistent execution.
• C. Packaging an application with its runtime, libraries, and configuration into a portable unit that shares the host operating system kernel is exactly what this concept provides.
• D. This concept describes a raw storage volume format, not a way to package an application and its dependencies for consistent execution.

Correct answer: A — They exclude rows containing a null value from the calculation, rather than treating the missing value as zero.

• A. This is correct because aggregate functions such as AVG typically ignore rows where the aggregated column is null, rather than treating the missing value as a zero.
• B. This is incorrect because substituting zero for a null would understate the average, and standard aggregate functions do not perform this substitution automatically.
• C. This is incorrect because encountering a null value in an aggregated column does not typically cause the query to fail or halt execution.
• D. This is incorrect because aggregate functions do not pre-fill null values with a computed average before running; they simply exclude the null rows from the calculation.

Correct answer: A — Decimal, because it stores an exact fixed-precision value, avoiding the rounding errors of floating-point storage.

• A. Decimal is correct because it stores numbers with exact fixed precision, which avoids the rounding imprecision that floating-point types can introduce in monetary calculations.
• B. Float is incorrect because it approximates values using binary floating-point representation, which can introduce small rounding errors unsuitable for exact monetary amounts.
• C. Integer is incorrect because it cannot represent fractional cent values on its own, forcing an awkward workaround that decimal handles natively.
• D. Boolean is incorrect because it can only represent two states, such as true or false, and cannot store a numeric monetary amount at all.

Correct answer: A — An application programming interface that returns structured exchange-rate data on each programmatic request

• A. An API lets the application request current exchange-rate data programmatically as often as needed, matching the automated, several-times-per-day requirement without manual steps.
• B. A manually emailed export depends on a person sending a file, which breaks the requirement for automated, on-demand refreshes multiple times a day.
• C. Scraping a webpage into a saved HTML file introduces manual or brittle parsing steps and was explicitly ruled out by the scenario.
• D. A database replica the developer cannot access directly is not a usable data source for this application, regardless of how current its data is.

Correct answer: A — A data lake, storing raw structured, semi-structured, and unstructured data without a fixed schema upfront

• A. A data lake accepts raw sensor readings, images, and unstructured notes in their native format without demanding a schema before ingestion, which matches this company's need to store first and decide analysis later.
• B. A data mart holds already-cleaned, department-scoped data prepared for reporting, which contradicts the need to keep raw, undecided-format data.
• C. A data warehouse requires incoming data to conform to a predefined relational schema, which raw images and unstructured notes cannot satisfy without prior transformation.
• D. A data silo describes an isolated store that other teams cannot access; it is an organizational problem rather than a repository designed for raw multi-format ingestion.

Correct answer: A — The data set contains all the records and field values that are expected to be present for its intended purpose

• A. This definition is correct because completeness measures whether the expected records and field values are actually present, which is the standard meaning of the term in data exploration.
• B. This definition is incorrect because the absence of duplicate rows describes deduplication, not completeness.
• C. This definition is incorrect because staying within a statistical range describes the absence of outliers, not whether expected data is present.
• D. This definition is incorrect because consistent units of measure describe a standardization or transformation step, not whether the data set has all its expected values.

Correct answer: A — Confirm with the source system owner whether `Retired` is a legitimate status that should be added to the accepted value list

• A. Confirming with the source system owner is correct because a validation rule that rejects a value not previously anticipated should be reviewed against real business meaning before deciding whether to expand the rule or correct the data.
• B. Automatically converting to `Terminated` without review is incorrect because it assumes equivalence between two distinct statuses without confirming that assumption is actually accurate for payroll purposes.
• C. Deleting the rows immediately is incorrect because it removes legitimate employee records before determining whether `Retired` should simply be added as a valid status.
• D. Ignoring the failed rows is incorrect because validation rules exist specifically to catch values that do not conform to expected business logic, and skipping the review defeats the purpose of running the check.

Correct answer: A — Concatenate the twelve files by appending their rows into a single result set, since the column structure is identical across all of them.

• A. Concatenating, also called a union, appends the rows of each file to the others when the column layout matches, producing one continuous table of all twelve months of sales rows, which is what the scenario requires.
• B. Joining on a shared column combines columns from two sources side by side for matching keys, which would widen the table with duplicate or mismatched columns rather than stack the same eleven columns across all twelve files.
• C. A nested query embeds one query inside another to filter or compute a value, but it does not stack independent files with identical structure into one combined row set.
• D. Grouping and aggregating collapses rows into summary totals per group, which would lose the individual sales records rather than preserve every row from all twelve files.

Correct answer: A — Surveying, since the analyst is directly asking customers structured questions to gather their opinions and feedback.

• A. Surveying is the practice of collecting information directly from people through structured questions, which matches the analyst designing and sending a questionnaire with rating-scale and open-text items to gather customer opinions.
• B. Sampling refers to selecting a representative subset of an existing dataset for analysis, not to designing and distributing new questions to gather fresh opinions from customers.
• C. Query optimization is a set of techniques for improving database query performance, such as indexing and parameterization, and has no relationship to designing a customer questionnaire.
• D. Data integration means combining data from multiple existing sources into a unified view, which is a separate later activity from the act of collecting new opinion data through a questionnaire.

Correct answer: A — Standardization, which rescales values based on the mean and standard deviation of the distribution so they can be compared across differing scales.

• A. Subtracting the mean and dividing by the standard deviation is the defining calculation of standardization, which puts differently distributed variables onto a common, comparable scale.
• B. Binning groups values into discrete performance categories and does not involve the mean-and-standard-deviation calculation used to make two distributions comparable.
• C. Imputation addresses gaps caused by missing scores and has no bearing on scores that are already present and simply need to be rescaled.
• D. Augmentation brings in attributes from an outside source, which is unrelated to a purely mathematical rescaling of an existing numeric column.

Correct answer: A — Merging joins datasets side by side using a shared key column, while appending stacks datasets with matching structure on top of each other.

• A. Merging aligns two datasets horizontally by matching a shared key, while appending stacks compatible datasets vertically, adding more rows rather than more columns.
• B. Neither operation inherently requires deleting duplicates or guarantees no rows are removed; duplicate handling is a separate decision made independently of whether the operation is a merge or an append.
• C. Merging can join on keys of any data type, including text identifiers, so the operation is not restricted to numeric columns alone.
• D. Neither merging nor appending is defined by changing a column's data type; type conversion is a separate transformation applied independently of combining datasets.

Correct answer: A — `IF`, a logical function that tests a condition and returns one value when true and another when false.

• A. Testing whether the order total exceeds $500 and returning a different result for true versus false matches exactly what this function is designed to do.
• B. Removing extra spaces from text has no ability to test a numeric threshold or return a conditional Yes or No result.
• C. Extracting the calendar year from a date value has no relationship to testing whether an order total exceeds a threshold.
• D. Calculating a mean across a range of values produces a single number, not a conditional Yes or No label for each order.

Correct answer: A — A string function such as `TRIM`, which removes leading and trailing spaces from a text value.

• A. Removing the extra leading and trailing spaces directly fixes the exact-match lookup problem described for these imported customer names.
• B. Requiring multiple conditions to be true evaluates logical conditions and does not remove extra spaces from imported text values.
• C. Adjusting decimal precision applies to numeric values and has no effect on extra spaces within a text field.
• D. Extracting a calendar year applies to date values and has no relationship to fixing extra spaces in a text name field.

Correct answer: A — First-contact resolution rate, targeted at 85% or higher, tracked weekly against the prior period.

• A. This is correct because it names a specific, measurable indicator (first-contact resolution rate), a target threshold, and a tracking cadence, all core traits of a well-formed KPI.
• B. Counting coffee breaks is measurable but has no established connection to support team performance or service goals, so it does not function as a meaningful KPI.
• C. A meeting room preference is a logistical detail, not a measurable performance indicator, so it cannot serve as a KPI for the team's weekly review.
• D. A vague statement to 'do better' lacks any specific, measurable indicator or target, which fails the basic definition of a KPI.

Correct answer: A — Add the non-aggregated column to the GROUP BY clause so every selected column is either grouped or aggregated.

• A. This specific error means a selected column must be grouped or aggregated, so adding it to the GROUP BY clause directly satisfies that requirement.
• B. This is a query-parsing error unrelated to the network session, so restarting the connection would not change the outcome.
• C. The error occurs regardless of the data's integrity, since it is about how the query is structured, not about corrupted values.
• D. GROUP BY behavior is standard SQL syntax, not a firmware limitation, so no vendor patch is needed to resolve this error.

Correct answer: A — Whether the source table's schema, column names, or data types changed as part of the upstream migration.

• A. A migration timed exactly with the failure, and no query changes on this end, strongly points to a schema or structure change made by the upstream system.
• B. Local disk space affects how a report renders on one machine, not why a server-side query would suddenly fail after an unrelated migration.
• C. Branding and accessibility of the report's visuals have no bearing on whether the query executing against the migrated source succeeds.
• D. A certification exam listing is unrelated to diagnosing a technical schema change caused by an infrastructure migration.

Correct answer: A — Add descriptive alt text and a written data summary for each chart so a screen reader can convey the chart's meaning aloud.

• A. Alt text and written summaries are correct because screen readers convert text to speech, so a chart without a text equivalent conveys nothing to a user relying on audio output.
• B. Increasing font size addresses low vision, a visual accessibility need, not the auditory access provided by screen-reader-compatible text descriptions.
• C. A colorblind-safe palette addresses a visual color-perception need and does not help a screen reader convey chart content through audio.
• D. A sharper, zoomable PDF improves visual clarity at higher resolution but does not add any text a screen reader could announce, so it does not address auditory access.

Correct answer: A — An ad hoc report, since it is generated for this single, unplanned occasion rather than on a repeating schedule

• A. A one-time request tied to a specific upcoming meeting, rather than an ongoing schedule, is the defining trait of an ad hoc report.
• B. The director explicitly does not want this produced every month, so scheduling it as a recurring report would contradict the stated requirement.
• C. Comparing two fixed prior periods does not require continuous refreshing, so a dynamic dashboard is not necessary to satisfy this one-time comparison.
• D. The director asked the analytics team to produce the report rather than build it themselves, so a self-service portal does not match the stated request.

Correct answer: A — A recurring report paired with snapshot versioning captured at a consistent weekly point in time

• A. Automatically sending the report every Monday without manual requests matches recurring frequency, and using data as of a fixed prior point (Sunday night) matches snapshot versioning.
• B. The scenario explicitly states the report goes out automatically every Monday rather than being requested, which rules out describing it as ad hoc.
• C. The report uses a fixed Sunday-night capture rather than data that changes every time it is viewed, so real-time versioning does not match the described behavior.
• D. The report is sent automatically on a schedule rather than only when requested, so labeling it ad hoc contradicts the stated weekly automation.

Correct answer: A — Apply the company's approved logo, color palette, and typeface to the report so it matches other official external materials

• A. Applying the approved logo, palette, and typeface is correct because consistent branding across external materials builds recognition and signals that the report is official company output.
• B. Branding affects how credible and recognizable a report appears to external stakeholders, so leaving default styling in place contradicts the style guide's purpose.
• C. Replacing charts with plain text avoids the branding question but sacrifices the clarity that visualizations provide, which is not what the style guide requires.
• D. Adding multiple unrelated logos creates visual confusion about which department or brand the report represents, rather than the single consistent identity the guide specifies.

Correct answer: A — Review the calculation logic and query steps that run during the refresh to find a newly introduced inefficient join or nested calculation

• A. A calc/code review of the refresh logic can surface a recently added inefficient join, a recalculated measure, or an unoptimized query step that is now taking longer even though data volume is stable, which directly targets the refresh-rate regression.
• B. Checking the color palette against branding guidelines is a visual design concern and has no bearing on how long the refresh query takes to execute.
• C. Reducing visible legends changes rendering load slightly but does not address a refresh that is slow because of the underlying query or calculation logic, which is the described symptom.
• D. Lowering delivery frequency hides the slow refresh from users less often but does not diagnose or fix why the refresh itself has gotten slower.

Correct answer: A — Restructure the source data with pre-aggregated, filterable summary tables so the load time and the filter's underlying field references are corrected at once.

• A. Restructuring the source into pre-aggregated summary tables with correct filterable fields reduces the row-level volume driving the load-time problem and fixes the store field the filter depends on, addressing both symptoms from their shared root cause.
• B. Adding a versioning toggle changes how current the displayed data is but still queries the same unfiltered, transaction-level volume, so it would not resolve the load time or the filter mismatch.
• C. Explaining the slowness to leadership communicates the problem but does not reduce data volume or fix the broken filter reference.
• D. Highlighting slow sections with color is a cosmetic indicator of the problem and does not reduce query volume or correct the filter's underlying issue.

Correct answer: A — The two axes were scaled independently, so aligning their ranges visually can overstate or understate how closely the two lines actually move together.

• A. Independently scaled dual axes are correct because choosing each axis range separately can visually align two unrelated lines, creating an impression of correlation that the underlying numbers may not support.
• B. The scenario describes an existing chart with two visible lines and axes, not a pivot table, so this option does not match what is actually being described.
• C. Color scheme choice affects how categories are distinguished visually but has no bearing on how the two numeric axis ranges were scaled relative to each other.
• D. A missing legend would make it harder to identify which line is which, but it does not change the independent axis scaling that is causing the misleading visual alignment.

Correct answer: A — Hierarchy structure

• A. Modeling categories, subcategories, and SKUs as nested parent-child levels is exactly what this concept documents, capturing how data elements relate to one another structurally.
• B. This concept traces where a data element originated and how it was transformed, not how categories and subcategories nest within one another.
• C. This concept identifies the single authoritative dataset for a given fact, which is unrelated to describing parent-child category relationships.
• D. This artifact defines field names, types, and valid values for a dataset, not the nested organization of categories and subcategories.

Correct answer: A — How frequently a dataset or report is updated with new data from its source, such as hourly, daily, or weekly.

• A. A refresh interval specifies the cadence at which a dataset or report pulls new data from its source, whether that cadence is hourly, daily, or weekly.
• B. This describes a snapshot, which captures a fixed historical copy, rather than the recurring cadence at which live data is refreshed.
• C. This describes access control, which governs who can view or modify data, and is unrelated to how often a dataset's contents are updated.
• D. This describes a query cache setting, which controls result reuse at the engine level, not the business cadence at which a dataset is refreshed with new source data.

Correct answer: A — Encrypt the connection between the source database and the cloud warehouse using TLS so the transferred records are unreadable if intercepted in transit.

• A. TLS encryption on the transfer channel is exactly what encryption in transit means: it protects data while it moves across a network, which is the exposure this nightly transfer creates. Anyone intercepting the traffic would only see ciphertext.
• B. Compression changes file size and transfer speed but does nothing to prevent an interceptor from reading the contents once decompressed. It is a performance technique, not a protection against exposure during transmission.
• C. Encrypting the source database volume protects data at rest on that server's disks, not data while it is moving across the network to the warehouse. The two controls address different stages of the data's lifecycle.
• D. Restricting who can query the warehouse after the load is an access control measure that governs use once the data has already arrived. It does not protect the records during the nightly transfer itself.

Correct answers: A, B — Anonymize the dataset by removing or generalizing fields, such as exact birth dates, that could be combined with outside data to re-identify a respondent.; Suppress or aggregate small population subgroups so that no combination of remaining fields narrows the data down to a single identifiable respondent.

• A. Removing or generalizing quasi-identifying fields like exact birth dates is a core anonymization step that prevents re-identification through combination with outside datasets, which directly matches the agency's stated goal.
• B. Suppressing or aggregating small subgroups prevents a respondent from being isolated through a rare combination of attributes, a standard anonymization technique for public statistical releases where no individual should be re-identifiable.
• C. Retaining the original identifying values elsewhere means the release is reversible in principle, which fails the goal of an anonymized public dataset where identity must never be recoverable by anyone, including the publisher.
• D. A public research repository is meant to be openly accessible; restricting it to the agency's own analysts contradicts the purpose of a public release and does not itself address re-identification risk within the published content.
• E. Distributing the decryption key to any requesting researcher effectively makes the file's contents as accessible as if it were unencrypted, and encryption does nothing to remove identifying detail from the dataset itself.

Correct answer: B — User acceptance testing, having the actual business users validate that the delivered report meets their working requirements

• A. Subjecting the query engine to a simulated user surge measures performance under load, not whether the report's content satisfies the marketing team's working needs.
• B. Having the actual business users run their typical filters and confirm the segments and totals meet their needs before sign-off is the defining activity of user acceptance testing.
• C. Checking version history confirms that changes were tracked and approved, which is unrelated to marketing staff validating the report's content against their needs.
• D. Scanning the underlying tables for completeness and uniqueness happens earlier, before the report is built, and does not involve business users validating the finished deliverable.

Correct answer: A — International Organization for Standardization (ISO), which publishes internationally recognized data quality management standards

• A. The International Organization for Standardization publishes internationally recognized standards that specifically address data quality management, making it the reference the manager is looking for.
• B. This organization's standards and certifications focus on project management methodologies, which is a different discipline than data quality management practices.
• C. This organization's technical standards are primarily aimed at electrical and electronics engineering rather than data quality management practices.
• D. This organization's frameworks are best known for cybersecurity and privacy guidance for U.S. federal agencies, not a general-purpose international data quality management standard.

Correct answer: A — Define retention periods per record type based on applicable regulations, then automatically archive or delete records once each period expires

• A. Sound retention practice ties specific retention periods to record type and the regulations governing that type, then enforces disposal or archival automatically once the period lapses.
• B. Keeping everything indefinitely in production ignores regulatory minimums and maximums, inflates storage cost, and increases exposure if a breach occurs, since more data sits at risk longer.
• C. Deleting records the moment treatment ends can violate legally mandated minimum retention periods that many healthcare regulations impose for patient records.
• D. Leaving retention to informal departmental judgment creates inconsistent compliance postures and makes it difficult to demonstrate adherence to a regulator during an audit.