Difficulty: Medium
When sales target is $1 million and actuals are $900,000, what is the shortfall percentage?
A 5% below target
B 100% below target
C 10% below target
D 10% above target
Answer: C. 10% below target. The shortfall is $100,000, which is 10% of the $1 million target.
Why are weighted averages better than simple averages for unequal groups?
A They sort data by date
B They remove missing values
C They prevent all outliers
D They reflect each group’s size
Answer: D. They reflect each group’s size. Weights account for the different sizes or denominators of groups.
What relationship is typical between a dimension and fact table?
A No key relationship at all
B One dimension to many facts
C One fact to many users
D Many dimensions to one chart
Answer: B. One dimension to many facts. A dimension key commonly appears in many related fact rows.
How can duplicate fact rows distort a KPI?
A They improve data lineage
B They inflate summarized values
C They change dimension labels only
D They prevent model refresh
Answer: B. They inflate summarized values. Duplicated facts can be counted twice and overstate totals.
Why can averaging group percentages misstate an overall rate?
A Group denominators differ
B Percentages cannot be compared
C The model needs a dashboard
D All groups have equal size
Answer: A. Group denominators differ. An overall rate should use combined numerators and denominators when group sizes differ.
How do dimensions support a star-schema report?
A Store report page layouts
B Filter and group fact data
C Run source-system transactions
D Replace every measure
Answer: B. Filter and group fact data. Dimension tables provide grouping and filtering context for fact measures.
Which method preserves previous customer category values?
A Overwrite the only value
B Rename the report page
C Delete past fact rows
D Create a new dimension version
Answer: D. Create a new dimension version. Type 2 history adds a new dimension record instead of overwriting the old one.
What does a filter context determine?
A Refresh capacity limits
B Data source credentials
C Rows included in a calculation
D Dimension storage format
Answer: C. Rows included in a calculation. Filter context controls which data rows contribute to a measure.
Why can many-to-many relationships complicate reporting?
A They eliminate dimension tables
B Filters may create ambiguous totals
C They remove all duplicates
D They always improve performance
Answer: B. Filters may create ambiguous totals. Many-to-many paths can create ambiguous filtering or double counting if modeled poorly.
Which security method limits rows by the viewer’s role?
A Report subscription
B Column formatting
C Row-level security
D Incremental refresh
Answer: C. Row-level security. RLS filters which records a user can view according to role rules.
Why reconcile a report total with its source?
A Confirm transformation and aggregation accuracy
B Increase dashboard speed
C Change historical grain
D Simplify visual colors
Answer: A. Confirm transformation and aggregation accuracy. Reconciliation checks whether data processing preserves expected totals.
When an unknown dimension key appears, what should a model do?
A Invent a customer name
B Use a governed unknown member
C Change the source amount
D Drop the fact silently
Answer: B. Use a governed unknown member. An unknown member can preserve fact rows while flagging missing dimension context.
What does a semantic model centralize?
A Relationships, measures, and definitions
B User training attendance
C Server replacement schedules
D Only dashboard screenshots
Answer: A. Relationships, measures, and definitions. Semantic models define shared data structures, relationships, and calculations for reports.
How can lineage help investigate a wrong KPI?
A Delete the measure owner
B Change the target value
C Trace sources and transformations
D Hide report filters
Answer: C. Trace sources and transformations. Lineage shows the upstream sources and processing behind a result.
Which refresh design suits a large historical table with a date key?
A Full reload after each visual
B Manual export only
C Incremental refresh by period
D Dimension overwrite each hour
Answer: C. Incremental refresh by period. Incremental refresh can process selected time ranges instead of all history.
Why certify a shared BI dataset?
A Signal governed, trusted reuse
B Remove ownership details
C Hide its refresh time
D Increase its row count
Answer: A. Signal governed, trusted reuse. Certification marks content that meets an organization’s trust and governance criteria.
What does an outlier represent in a dataset?
A Missing dimension definition
B A report subscription
C A valid relationship key
D Unusually distant observation
Answer: D. Unusually distant observation. Outliers are observations that differ substantially from typical values.
How should a highly skewed distribution be summarized robustly?
A Use the median
B Sum the category labels
C Use the maximum only
D Count the report pages
Answer: A. Use the median. The median is less affected by extreme values than the mean.
Why can correlation in a dashboard mislead decision makers?
A Dimensions are always inaccurate
B Measures cannot be filtered
C Association does not prove causation
D Charts never show trends
Answer: C. Association does not prove causation. Correlation describes association but does not establish a causal relationship.
What is a leading indicator intended to show?
A Warehouse table size
B Only final historical outcomes
C A dashboard refresh status
D Signals about future performance
Answer: D. Signals about future performance. Leading indicators provide signals that may precede later outcomes.
Which calculation reports percentage variance from target?
A Actual less target over target
B Target divided by actual only
C Actual plus target
D Target minus actual, times 100
Answer: A. Actual less target over target. Percent variance typically divides the difference by the target or baseline.
How can a BI team detect stale reporting data?
A Increase row-level security
B Monitor refresh time and freshness
C Rename the semantic model
D Add more chart colors
Answer: B. Monitor refresh time and freshness. Freshness monitoring compares update timestamps with the expected refresh cadence.
Which comparison uses comparable time periods?
A Current week versus full year
B Revenue versus number of pages
C Actual sales versus customer count
D Current quarter versus prior quarter
Answer: D. Current quarter versus prior quarter. Comparable periods support more meaningful trend and performance interpretation.
What does aggregation do to detailed records?
A Changes source ownership
B Adds new raw transactions
C Creates a report subscription
D Summarizes at a higher level
Answer: D. Summarizes at a higher level. Aggregation combines detailed data into summaries such as totals by month.
Why standardize KPI definitions across teams?
A Keep results comparable and trusted
B Avoid documenting ownership
C Remove dimension analysis
D Make every metric identical
Answer: A. Keep results comparable and trusted. Shared definitions prevent teams from interpreting the same KPI differently.