Home / Data Analyst practice test / Working with Dashboards and Visualizations in Databricks

Free · 8 questions with explanations

Working with Dashboards and Visualizations in Databricks: Databricks Data Analyst Associate Practice Questions

Exam-style questions on Working with Dashboards and Visualizations in Databricks. Pick your answer, then open the explanation to see why it's right — and why the other options are wrong.

Take a full timed test free → All Data Analyst topics
1 Working with Dashboards and Visualizations in Databricks

After a scheduled AI/BI dashboard refresh completes, what does a subscribed Slack channel receive as the notification delivery?

  1. AAn HTML email with embedded visualization screenshots sent to the Slack channel's associated workspace email address, identical to what individual email subscribers receive in their inboxes
  2. BOnly a direct hyperlink to open the published dashboard in Databricks, with no visual snapshot attached, because Slack's third-party integration model does not permit image or file attachments from external services
  3. CA PNG image snapshot visible directly in the Slack channel, plus a direct link to open the dashboard in Databricks, and a PDF attachment included in the Slack message thread
  4. DA CSV file attachment containing the raw data from all dashboard datasets, enabling Slack channel members to perform their own analysis without requiring a Databricks workspace account
  5. EA JSON export of all widget configurations and dataset result sets, allowing Slack users to reconstruct the dashboard view locally in any compatible business intelligence application
Show answer & explanation

Correct answer: C

WHY C: According to Databricks documentation, Slack channel subscribers receive: (1) a PNG image snapshot visible directly in the channel, (2) a direct link to open the dashboard in Databricks, and (3) a PDF attachment in the message thread. WHY NOT A: Email subscribers receive a PDF snapshot delivered to their email inbox; Slack subscribers receive a PNG image + link + PDF thread attachment — not an HTML email sent to a Slack inbox. WHY NOT B: Slack integrations configured by workspace admins do support image and file attachments from Databricks; a link-only notification is incorrect. WHY NOT D: No CSV data attachments are sent via dashboard subscriptions; subscriptions deliver visual snapshots (PNG/PDF), not raw data exports for external analysis. WHY NOT E: JSON configuration exports are used in the dashboard import/export workflow for transferring dashboards between workspaces; subscription notifications deliver visual snapshots, not JSON reconstruction files.

2 Working with Dashboards and Visualizations in Databricks

A data analyst needs to configure an AI/BI dashboard to refresh automatically every morning at 6:00 AM in their local time zone. Which steps correctly describe the process using the dashboard UI?

  1. AOpen the published dashboard, click the 'Auto-refresh' toggle near the upper toolbar, set the interval dropdown to 'Daily', type '06:00' in the time field, and confirm; no further warehouse selection is needed
  2. BOpen the draft dashboard's Data tab, click 'Add trigger', set the cron expression to 0 6 * * *, select the compute cluster, and click 'Enable' to activate the recurring dataset refresh
  3. CIn the dashboard UI, click Schedule and then + Schedule, set the frequency and start time, select the Timezone, optionally choose a SQL warehouse for the scheduled run, and click Create
  4. DOpen the associated Lakeflow Jobs page, add a new task with type 'Dashboard refresh', configure the trigger to run daily at 6:00 AM, and deploy the job pipeline to run on a scheduled cluster
  5. ENavigate to the SQL Warehouse settings page, select the warehouse used by the dashboard, enable 'Auto-refresh connected dashboards', and set the daily refresh time to 06:00 in the warehouse configuration
Show answer & explanation

Correct answer: C

WHY C: To create a schedule on an AI/BI dashboard, the author clicks Schedule → + Schedule in the upper-right corner of the dashboard. The scheduling dialog allows setting the refresh frequency, start time, time zone, filter values (default or current), and an optional SQL warehouse override. Clicking Create saves the schedule. WHY NOT A: There is no 'Auto-refresh toggle' in the dashboard toolbar; schedule creation goes through the Schedule button and the scheduling dialog modal, not a simple toggle. WHY NOT B: Scheduled refreshes are not configured from the dataset Data tab with cron triggers; the scheduling feature is accessed from the Schedule button in the main dashboard UI. WHY NOT D: While Lakeflow Jobs can be used for advanced orchestration that includes dashboard refreshes, the objective refers to the built-in UI-driven scheduling feature, which is accessed directly from the dashboard's Schedule button. WHY NOT E: SQL Warehouse settings do not include an 'Auto-refresh connected dashboards' setting; schedules are configured per-dashboard, not per-warehouse.

3 Working with Dashboards and Visualizations in Databricks

After running a SQL query in the Databricks SQL editor, a data analyst wants to create a bar chart from the result set. What is the correct built-in way to create the chart without leaving the SQL editor?

  1. AExport the query result to a CSV, open it in a local spreadsheet application, create the bar chart there, and then upload it back as an image widget on an AI/BI dashboard canvas
  2. BClick the '+' tab next to the Results tab in the SQL editor to add a new visualization, then configure the chart type and column mappings in the visualization editor panel
  3. CCopy the result set into a Python notebook cell, convert it to a Spark DataFrame, call display(df) on it, and configure the resulting interactive chart from the display() output area
  4. DNavigate to the AI/BI Dashboards page, create a new dashboard, paste the SQL statement as a new dataset, and then add and configure a visualization widget on the dashboard canvas
  5. EAdd a PLOT TYPE='bar' XAXIS=col1 YAXIS=col2 clause at the end of the SQL query before re-running it, which causes Databricks to render the bar chart automatically below the results grid
Show answer & explanation

Correct answer: B

WHY B: In the Databricks SQL editor, after running a query, you click the '+' tab next to the Results tab to create a new visualization directly from the result set. The visualization editor lets you choose the chart type (bar, line, scatter, etc.) and configure column mappings — all without leaving the SQL editor. WHY NOT A: Exporting to CSV and using external tools is a valid workaround but is not the built-in SQL editor workflow for creating visualizations. WHY NOT C: While display() works in a separate Python notebook, it requires a different environment and converting the result set to a DataFrame; this bypasses the SQL editor's native visualization feature. WHY NOT D: Navigating to AI/BI Dashboards is a separate product surface and is not required for creating a quick visualization from the SQL editor. WHY NOT E: There is no PLOT keyword in Databricks SQL syntax; chart creation uses the UI editor, not SQL-level keywords.

4 Working with Dashboards and Visualizations in Databricks

A dashboard author enables 'Allow multiple selections' on a parameter named ':product_types'. Which SQL WHERE clause pattern must the parameterized dataset query use for multi-select to work correctly, including handling the 'All' selection?

  1. AWHERE product_type IN (:product_types), because Databricks automatically expands the array parameter into a comma-separated IN-list at query runtime when multiple selections are active
  2. BWHERE product_type = :product_types, because enabling multiple selections causes Databricks to implicitly loop over the array and perform OR comparisons against each element at runtime
  3. CWHERE ARRAY_CONTAINS(:product_types, product_type) OR :product_types IS NULL — uses ARRAY_CONTAINS for multi-value matching and IS NULL to handle the 'All' option
  4. DWHERE product_type LIKE :product_types with a '%' delimiter, because Databricks pattern-matching syntax is the supported mechanism for matching multiple simultaneous parameter values
  5. EWHERE SPLIT(:product_types, ',') = product_type, which splits a comma-delimited parameter string into an array and compares each element against the product_type column value for a match
Show answer & explanation

Correct answer: C

WHY C: When 'Allow multiple selections' is enabled, the parameter is inserted into the query as an array. The ARRAY_CONTAINS function is required to check whether product_type belongs to the parameter array, and OR :product_types IS NULL handles the 'All' scenario by returning all rows when no selection is made (parameter is NULL). WHY NOT A: The IN (:param) syntax does not work for multi-select array parameters in Databricks; Databricks does not auto-expand array parameters into IN-list syntax. ARRAY_CONTAINS is required. WHY NOT B: Using = with a multi-select parameter compares the entire array object against a scalar column value, causing a type mismatch or incorrect results rather than OR comparisons. WHY NOT D: LIKE pattern-matching is valid for single-value text filters but cannot handle array parameters produced by multi-select; there is no Databricks-supported multi-value LIKE syntax for parameters. WHY NOT E: SPLIT is used to split delimited strings; multi-select parameters are inserted as arrays directly by Databricks, not as comma-delimited strings requiring post-processing with SPLIT.

5 Working with Dashboards and Visualizations in Databricks

A dashboard author publishes a dashboard and wants viewers who do NOT have direct access to the underlying Unity Catalog tables to still see the full dashboard data. Which publishing data permission setting achieves this?

  1. AIndividual data permissions, which grants each viewer temporary elevated credentials to the underlying Unity Catalog tables for the duration of their current browser session
  2. BShared data permissions (the default), where all viewers run queries using the publisher's credentials, allowing them to see data even if they have no direct access to the underlying tables
  3. CConsumer access entitlement mode, which bypasses Unity Catalog row-level security and grants any account-registered viewer read access to all tables used by the dashboard datasets
  4. DWorkspace-level open permissions configured by the workspace admin, which temporarily grants CAN USE on all catalogs and schemas referenced by the dashboard's dataset definitions
  5. EEmbedded data permissions, which package the most recent query result snapshot directly into the published dashboard so that viewers query the cached snapshot rather than live catalog tables
Show answer & explanation

Correct answer: B

WHY B: Shared data permissions (the default) cause all viewers to run queries using the publisher's credentials. This allows users who lack direct access to the underlying data or compute to still view the full dashboard. This is explicitly described as the option to 'use when sharing with users who do not have access to the underlying data.' WHY NOT A: Individual data permissions have the opposite effect — viewers use their own credentials, so users without table access would see no data or receive permission errors. WHY NOT C: There is no 'Consumer access entitlement mode' as a publishing option; Consumer access is a workspace entitlement limiting users to read-and-run-only on the dashboard UI, not a bypass of Unity Catalog security. WHY NOT D: There is no workspace-level open permissions setting for catalogs triggered by dashboard publishing; catalog and schema permissions are managed separately in Unity Catalog. WHY NOT E: There is no 'Embedded data permissions' option in the publishing dialog; dashboards always query live data (cached or freshly run), not pre-packaged snapshots.

6 Working with Dashboards and Visualizations in Databricks

A data analyst is monitoring daily active users across a 12-month period and wants to display the overall trend, including seasonal spikes, as clearly as possible. Which visualization type is most effective for this use case?

  1. AA pie chart that partitions the 12-month total active users into proportional slices by month, making the relative contribution of each month visually apparent at a single glance
  2. BA scatter plot that plots each day's active user count as an individual data point on a two-axis chart, enabling the viewer to explore correlation between elapsed time and user activity volume
  3. CA line chart that connects data points along a continuous time axis, making the overall direction, rate of change, and seasonal spikes in daily active users easy to identify
  4. DA grouped bar chart that renders one bar per month with its height representing the total active users, allowing viewers to compare monthly magnitudes side-by-side without a connected trend line
  5. EA heat map that represents user activity intensity as a color gradient on a two-dimensional grid of days and hours, optimized for identifying recurring intra-week or intra-day cyclical patterns
Show answer & explanation

Correct answer: C

WHY C: A line chart is the most effective visualization for showing trends over time. It connects chronological data points into a continuous line, making it easy to see the overall direction, rate of change, and seasonal spikes in daily active users across the full 12-month period. WHY NOT A: A pie chart shows proportions of a whole at a single point in time and cannot effectively display trends, direction of change, or seasonal spikes across a continuous 12-month timeline. WHY NOT B: A scatter plot shows correlation or distribution between two numeric variables. While it can plot time-series points, it does not connect them to show trend direction, making seasonal spikes harder to identify than with a line chart. WHY NOT D: A grouped or stacked bar chart can compare monthly totals in magnitude, but it does not emphasize the continuous trend direction or seasonal spike patterns as clearly as a line chart does for time-series data. WHY NOT E: A heat map is useful for identifying recurring cyclical patterns (e.g., day-of-week × hour-of-day), but it is not the most effective choice for visualizing an overall 12-month directional trend with highlighted seasonal spikes.

7 Working with Dashboards and Visualizations in Databricks

A Databricks account user who has NOT been added to the workspace is granted CAN EDIT permission on a published dashboard through the Sharing dialog. What is this user's actual effective permission level on the dashboard?

  1. ACAN EDIT as specified, because the Sharing dialog overrides workspace membership requirements and enforces the assigned permission level regardless of whether the user has workspace access
  2. BCAN MANAGE, because the system automatically elevates non-workspace users to the highest permission tier so they can administer the shared dashboard from the account level
  3. CCAN RUN (view-only), because non-workspace users are capped at CAN RUN regardless of any higher level assigned in the Sharing dialog, until workspace access is granted
  4. DNo access at all, because the Sharing dialog prevents elevated permissions from being assigned to users who are not yet workspace members and requires workspace membership before any access is granted
  5. ECAN VIEW read-only mode, which is a special intermediate permission tier automatically applied to account users without workspace access, falling between CAN RUN and CAN EDIT
Show answer & explanation

Correct answer: C

WHY C: According to Databricks documentation, users who do not have access to the workspace are limited to CAN RUN permissions. Even if CAN EDIT or higher appears in the Sharing dialog, those elevated permissions are not enforced until the user is added to the workspace. If the user is later added, the assigned permission level takes effect. WHY NOT A: The Sharing dialog does display the assigned permission, but it does not override the workspace membership requirement for elevated permissions; the actual effective permission is CAN RUN. WHY NOT B: Non-workspace users are not elevated to CAN MANAGE; they are capped at CAN RUN, the minimum permission tier for dashboard access. WHY NOT D: The Sharing dialog does allow higher permissions to be assigned to non-workspace users, and the assignments are retained; the elevated permissions simply are not enforced until workspace access is granted. WHY NOT E: CAN VIEW is not a distinct permission tier between CAN RUN and CAN EDIT in the dashboard ACL model; the tier applied to non-workspace account users is CAN RUN.

8 Working with Dashboards and Visualizations in Databricks

A data analyst wants to build an AI/BI dashboard that includes a free-form explanatory paragraph alongside a chart visualization. Which dashboard widget type is designed for adding narrative or Markdown-formatted text to the canvas?

  1. AA visualization widget configured with a 'text display mode' that renders plain text directly inside the chart container alongside the dataset output
  2. BA text widget, which allows dashboard authors to add free-form narrative or Markdown-formatted descriptive content beside other widgets on the canvas
  3. CA filter widget configured with a static string default value, which displays its label as descriptive text visible to viewers on the published dashboard canvas
  4. DA dataset widget that queries a table containing the paragraph text and renders it as a single-row, single-column table result displayed inline on the canvas
  5. EA metrics widget that renders an auto-generated text summary of all visualization outputs currently loaded and displayed in the active dashboard session
Show answer & explanation

Correct answer: B

WHY B: AI/BI dashboards support three widget types: visualization widgets (charts and tables from datasets), text widgets (for free-form or Markdown narrative), and image widgets. A text widget is the correct choice for adding explanatory paragraphs alongside charts. WHY NOT A: Visualization widgets render charts/tables from datasets and do not have a 'text display mode' for free-form narrative. WHY NOT C: Filter widgets expose interactive parameter controls to viewers; they are not intended to display static descriptive content as narrative text. WHY NOT D: Dataset widgets are not a standard canvas element type; datasets power visualizations but are not placed as text renderers on the canvas. WHY NOT E: There is no 'metrics widget' type in AI/BI dashboards; the three types are visualization, text, and image widgets.

Take the full Data Analyst practice test →

Keep going

Other Data Analyst topics


Executing queries using Databricks SQL and Databricks SQL Warehouses 8 practice questions → Analyzing Queries 8 practice questions → Developing, Sharing, and Maintaining AI/BI Genie spaces 7 practice questions → Managing Data 5 practice questions → Securing Data 5 practice questions → Understanding of Databricks Data Intelligence Platform 5 practice questions → Data Modeling with Databricks SQL 4 practice questions → Importing Data 4 practice questions →