Visualization Selection and Effectiveness Questions

Demonstrating the ability to choose appropriate chart types for different data patterns (trends over time, categorical comparisons, distributions, correlations). Creating visualizations that communicate clearly without ambiguity. Using color, formatting, and labels effectively to enhance understanding.

HardTechnical

38 practiced

How would you measure visualization effectiveness after launching a redesigned dashboard? Propose 3-4 quantitative and qualitative metrics (e.g., task completion time, error rate, NPS, heatmap clicks) and a brief experiment plan to A/B test the design.

HardTechnical

25 practiced

You need to present survival-like metrics (e.g., time-to-retention drop) in a dashboard where many customers are 'censored' (still active). What visualization(s) are appropriate (Kaplan–Meier curve, life table, cohort heatmap), and how would you explain censored observations to non-technical stakeholders?

MediumTechnical

20 practiced

Write a Power BI DAX measure (describe the logic, not necessarily full syntax) to compute a rolling 12-month revenue sum excluding the current partial month (i.e., include months M-12 to M-1). Explain filter context considerations and performance implications.

Sample Answer

Approach (logic):- For each visual row (current date context M), compute sum of Revenue for the 12 full months prior to the current month: months M-12 through M-1. Exclude any days of the current month so partial-month revenue is not included.

Sample DAX measure (conceptual):

dax

Rolling12FullMonthsRevenue =
VAR CurrentMonthStart = STARTOFMONTH( MAX( Dates[Date] ) )
VAR WindowStart = EDATE( CurrentMonthStart, -12 )  // first day of M-12
VAR WindowEnd = EDATE( CurrentMonthStart, -1 )     // last day of M-1 (ENDOFMONTH(WindowEnd) if needed)
RETURN
CALCULATE(
    SUM( Fact[Revenue] ),
    FILTER(
        ALL( Dates ),
        Dates[Date] >= WindowStart
        && Dates[Date] < CurrentMonthStart
    )
)

Filter context considerations:- Use ALL(Dates) (or REMOVEFILTERS) inside CALCULATE to override row-level filters so you can apply an explicit date window. But preserve other filters (product, region) by placing only Dates table in ALL.- MAX(Dates[Date]) reads the current report context; if the visual is by Month, ensure the Dates table has contiguous month granularity and that the model relates Dates -> Fact with a single active relationship.- If visuals use slicers with relative date filters, the measure logic should still derive CurrentMonthStart from context, but be careful when multiple months selected — define whether you want to aggregate per selection or use LASTDATE/MAX semantics.

Performance implications & best practices:- Prefer date-key integer comparisons (Dates[DateKey]) or use precomputed month keys to avoid heavy row-by-row FILTER over large Dates table.- If dataset is large, avoid FILTER(ALL(Dates), ...) scanning many rows each evaluation. Use DATESBETWEEN or built-in time-intelligence like DATEADD with a proper contiguous Date table which can be more optimized: Rolling12FullMonthsRevenue = CALCULATE( SUM(Fact[Revenue]), DATESBETWEEN( Dates[Date], WindowStart, EDATE(CurrentMonthStart,-1) ) )- Ensure Dates table is marked as Date Table and has continuous dates — this enables engine optimizations.- If used across many visuals, consider precalculating monthly aggregates in a summarized table to reduce measure runtime.

Edge cases:- Handle scenarios when there is less than 12 months of history (use conditional logic to sum available months).- Time zones/transaction timestamps: ensure Dates uses report-local date.

MediumTechnical

24 practiced

Write a SQL query (ANSI SQL) to prepare daily revenue for a time-series chart from orders(order_id, order_date, revenue). Ensure days with zero revenue appear in the result across a specified date range (2024-01-01 to 2024-03-31). Show sample result columns: date, daily_revenue.

Sample Answer

Approach: build a calendar (date) series for the requested range (using a recursive CTE in ANSI SQL), left-join aggregated orders by day, and coalesce NULLs to 0 so days with no revenue appear.

sql

WITH calendar AS (
  -- generate dates from start_date to end_date (ANSI-compatible recursive CTE)
  SELECT DATE '2024-01-01' AS dt
  UNION ALL
  SELECT dt + INTERVAL '1' DAY
  FROM calendar
  WHERE dt + INTERVAL '1' DAY <= DATE '2024-03-31'
),
daily_orders AS (
  -- aggregate revenue by order_date
  SELECT
    CAST(order_date AS DATE) AS dt,
    SUM(revenue) AS daily_revenue
  FROM orders
  WHERE order_date >= DATE '2024-01-01' AND order_date < DATE '2024-04-01'
  GROUP BY CAST(order_date AS DATE)
)
SELECT
  c.dt AS date,
  COALESCE(d.daily_revenue, 0) AS daily_revenue
FROM calendar c
LEFT JOIN daily_orders d ON c.dt = d.dt
ORDER BY c.dt
OPTION (MAXRECURSION 0); -- optional: for SQL Server; ignored in other engines

Key points:- The recursive CTE is ANSI SQL compatible; some engines provide functions (generate_series) you can use instead.- Coalesce ensures zero for dates with no orders.- Filter orders by date range before aggregation for performance.Edge cases:- timezone-aware timestamps — cast to DATE to normalize.- very large ranges — iterative generation can be costly; consider a numbers table for production.

MediumTechnical

25 practiced

Explain how axis scaling and baseline choices (starting at zero vs dynamic axis) affect bar charts and line charts. Provide examples where starting at zero is required and where a truncated axis might be acceptable with clear labeling.

Unlock Full Question Bank

Get access to hundreds of Visualization Selection and Effectiveness interview questions and detailed answers.

Join thousands of developers preparing for their dream job.