Data visualisation questions test whether you can pick the right chart and tell an honest, clear story with numbers.
Below are 60 Data Visualisation questions with the correct answer and a short explanation, grouped from fresher to experienced level. Try each one before you open the answer.
Which chart best shows how monthly revenue changed over the last two years?
Answer: A. Line chart
A line chart with time on the x-axis shows trend and seasonality clearly. Pie charts show parts of a whole at one moment, and histograms show a distribution, not change over time.
You need to compare total sales across 12 product categories so a manager can see the biggest and smallest. Which chart works best?
Answer: D. A bar chart sorted from largest to smallest
People compare lengths on a common baseline far more accurately than angles or areas. A sorted bar chart makes ranking obvious even with 12 categories. A pie chart with 12 slices is hard to read, and a line chart suggests an order or time trend that categories do not have.
You have ad spend and sales for 60 cities and want to see whether higher spend goes with higher sales. Which chart fits best?
Answer: B. A scatter plot of spend vs sales
A scatter plot puts one measure on each axis, so the shape of the cloud shows whether they move together, plus outliers. A line across cities implies an order that does not exist, and a pie shows only shares.
You need to rank 15 products by revenue, and the product names are long. Which chart is easiest to read?
Answer: C. A horizontal bar chart sorted by revenue
Horizontal bars leave room for long labels to be read normally, and sorting by value makes the ranking clear at once. Tilted labels are hard to read, and a donut with 15 slices makes comparison nearly impossible.
A line chart shows monthly sales for 12 regions on one set of axes, and the lines tangle into a "spaghetti" mess. What usually works better?
Answer: A. Small multiples, one panel per region
Small multiples give each region its own panel with the same scales, so each trend is clear and panels can still be compared. Another option is to grey out all lines and highlight the one or two that matter.
You want a compact view of sales across 8 categories and about 60 sub-categories, showing which areas are large. Which chart is designed for this?
Answer: B. A treemap
A treemap nests rectangles for sub-categories inside categories, with area showing size, so big areas stand out within the hierarchy. It is weaker for exact comparison, so pair it with a bar chart if precise ranking matters.
A chart has 4 coloured lines and a legend at the bottom. Readers keep looking back and forth. What is a simple improvement?
Answer: A. Label each line directly at its end
Putting the series name right next to the end of each line removes the need to decode colours through a legend. It speeds reading and also helps people who have trouble telling colours apart.
Your story is about the UAE region, but the chart shows 8 regions in 8 bright colours. What makes the message clearer?
Answer: B. Show UAE in colour and the rest in grey
Colour is a pre-attentive signal: the eye goes to the one coloured element before reading anything. Greying the rest keeps context while pointing straight at UAE. Many bright colours compete and say nothing about what matters.
A dashboard for Indian regional managers shows revenue as 125000000. How should it usually be displayed for them?
Answer: A. ₹12.5 Cr (or ₹12,50,00,000)
Indian readers think in lakhs and crores, so ₹12.5 Cr or the Indian grouping 12,50,00,000 is quickest to grasp. Raw digits and scientific notation are hard to read. Use one convention consistently across the dashboard.
Users want to right-click a store in a summary chart and jump to a page showing only that store's details. Which feature does this?
Answer: A. Drill-through
A drill-through page has the store field as its drill-through filter. Right-clicking a store elsewhere offers "Drill through", which opens that page filtered to the chosen store, with a back button to return.
In Tableau, Order Date on Columns shows as a green pill and draws a continuous time axis. What does a blue pill mean instead?
Answer: C. The field is discrete, giving separate headers
Blue pills are discrete: they create headers, one per value, such as separate month labels. Green pills are continuous: they create an axis. Dates can be used either way, which changes how the chart is drawn.
Users want to type a discount % and see projected revenue update in a Tableau view. Which feature allows a user-chosen value in a calculation?
Answer: C. A parameter
A parameter is a single user-controlled value, shown as a slider, list or input box, that calculated fields can reference. Here a calculation like [Revenue] * (1 - [Discount %]) updates as the user changes it.
On a Tableau dashboard, clicking a state on the map should filter the order table below to that state. What do you add?
Answer: B. A filter action on the dashboard
Dashboard filter actions pass the selected marks' values from a source sheet to target sheets. Choosing the map as source and the table as target filters the table to the clicked state.
A Tableau city field has both "Bombay" and "Mumbai" for the same city. Without changing the source, how do you show them as one?
Answer: D. Create a group combining them
A group combines several members of a dimension into one, so both spellings show as Mumbai. Sets define in/out membership for analysis, and hiding drops data. Fixing the source data is the long-term fix.
A business analyst is asked to "build a sales dashboard". What should they clarify first?
Answer: B. Who uses it and which decisions it supports
A dashboard is useful only if it answers the questions its users act on. Knowing the audience, their decisions and how often they look decides the metrics, level of detail and refresh rate. Colours and chart types come later.
A warehouse shift supervisor needs to watch orders waiting to be packed and act within minutes. What kind of dashboard is this?
Answer: B. An operational dashboard with near real-time data
Operational dashboards monitor current activity and trigger action, so they need fresh data, clear thresholds and alerts. Strategic dashboards track longer-term KPIs for leaders and can refresh daily or less often.
For a left-to-right reading audience, where should the most important KPIs usually go on a dashboard?
Answer: C. Top left, where the eye starts
Readers of left-to-right languages scan from the top left, often in an F or Z pattern, so key numbers belong there, followed by supporting detail lower down. For Arabic dashboards read right to left, mirror the layout.
Managers keep asking whether a dashboard's numbers are current. What simple addition helps most?
Answer: A. A "data as of" time near the title
Showing when the data was last refreshed, and the period it covers, tells readers how much to trust it and avoids decisions on stale numbers. Some teams also show a warning when the last refresh failed.
On one dashboard, UPI is blue in the first chart and orange in the second, while cards are blue in the second. What should change?
Answer: B. Keep one colour per category across all charts
Readers learn colour mappings quickly and carry them from chart to chart. Changing them forces re-reading legends and causes wrong conclusions. Fix the colour for each category across the dashboard, and use the same order too.
In a standard box plot of delivery times, what does the box itself (not the whiskers) cover?
Answer: A. The middle 50% of values, from Q1 to Q3
The box spans the first to third quartile, so it holds the middle half of the data, with a line at the median. Whiskers usually extend to 1.5 × IQR, and points beyond them are drawn as outliers.
A bar chart compares two months of sales: 980 and 1,020. The y-axis starts at 950, so the second bar looks more than twice as tall. What is the problem?
Answer: D. Bar length should be proportional to the value, so a bar chart y-axis should start at zero
Readers judge bars by length, so cutting the axis makes a 4% difference look like a 100% difference. Bar charts should start at zero. If small changes matter, use a line or dot chart with a clearly labelled axis, or show the percent change directly.
Daily sign-ups jump every weekend and dip every Monday. Why would you add a 7-day moving average line to the chart?
Answer: A. It removes the weekly pattern so the underlying trend is easier to see
Each point of a 7-day average covers exactly one of each weekday, so the weekly cycle cancels out and the trend shows. It does not make data more accurate and it lags behind sudden changes by a few days. Keep the raw line faintly in the background so spikes are still visible.
You have counts for 25 complaint types and want to show that a few types cause most complaints. Which chart is designed for that?
Answer: A. A Pareto chart: sorted bars, cumulative %
A Pareto chart sorts categories from largest to smallest and adds a cumulative percentage line, so you can read off, for example, that 5 types make up 80% of complaints. That tells the team where to focus first.
The question is "Has the share of UPI in payments grown each month?", while total payments also grew a lot. Which chart answers that directly?
Answer: A. A 100% stacked column of payment methods per month
A 100% stacked chart scales every month to the same height, so you see how the mix changes regardless of total growth. An absolute stacked chart mixes the share change with total growth, making the share hard to read.
A bar chart compares users across cities that range from 2,000 to 20,00,000. Small cities look like zero. When does a log scale help?
Answer: B. To compare relative gaps over a wide range
A log axis makes each step a fixed ratio, such as ×10, so small and large values both stay visible and equal percentage changes look equal. Label the axis clearly, because many readers assume a linear scale. Log scales cannot show zero or negatives.
A scatter of 3 lakh transactions (amount vs time of day) is a solid mass of dots, so you cannot see where most points are. What helps?
Answer: C. Use transparency or a hexbin density view
Overplotting hides density because dots cover each other. Semi-transparent points make dense areas darker, and hexbin or 2D histogram charts count points per cell. Sampling is another option if exact outliers do not matter.
You want to compare growth of Mumbai (₹500 Cr sales) and Kochi (₹20 Cr) since January. On a normal line chart Kochi looks flat. What helps?
Answer: D. Index both series to 100 at January
Indexing divides each series by its starting value and multiplies by 100, so both start at 100 and the lines show relative growth. A reading of 130 means 30% growth, whatever the city's size.
A daily sales chart compares October 2026 with October 2025, and shows a huge gap. Diwali fell in a different week each year. How do you make the comparison fair?
Answer: C. Align days around the festival date itself
Calendar dates hide that the big shopping days moved. Aligning by days before and after Diwali (or Eid, Ramadan and so on) compares like with like. Weekday alignment also matters, since weekends sell differently.
A line chart shows daily average delivery time. The service target is 30 minutes. How do you show the target most clearly?
Answer: D. Add a labelled reference line at 30 min
A reference line shows the target on the same scale as the data, so readers instantly see days above and below it. Label it directly ("Target 30 min"). A separate axis would make comparing the two harder.
A Power Query step added halfway through made refresh pull all 50 million rows from SQL Server instead of filtering at the source. What concept explains this?
Answer: D. Query folding stopped at that step
Query folding means Power Query turns steps into a single native query, so filtering happens in the database. Some steps cannot fold, and every step after them runs in Power Query on full data. Put foldable steps like filters first and check with "View Native Query".
Clicking a bar in "Sales by region" filters every visual on the page, but the "Total company revenue" card should never change. What do you set?
Answer: D. Edit interactions: set None for the card
Edit interactions lets you choose, for each source visual, whether other visuals filter, highlight or ignore its selection. Setting None for the card keeps it at the company total.
A Tableau text table has Years down the rows and Months across the columns. A RUNNING_SUM adds up years instead of months within each year. What do you change?
Answer: D. Set Compute Using to Table (across)
Table calculations run along the direction set in Compute Using. Table (down) adds year by year; Table (across) adds months left to right within each year row. Using specific dimensions makes the intent explicit and robust to layout changes.
A Tableau sheet has a Top 10 products filter and a Region filter set to UAE. Only 4 products appear. Why, and what is the fix?
Answer: B. Top N runs first; make Region a context filter
Tableau applies Top N filters before regular dimension filters, so it picks the overall top 10 and then keeps the 4 sold in the UAE. Adding Region to context makes it apply first, so Top N finds the top 10 within the UAE.
You have 60 regional managers and do not want 60 RLS roles. What is the common design for dynamic RLS?
Answer: A. One role: user-region table + USERPRINCIPALNAME()
A security table maps user emails to regions and is related to the region dimension. One role with a rule like [Email] = USERPRINCIPALNAME() filters it, and the filter flows to the facts. Adding a manager just means adding a row.
A designer uses light grey (#BBBBBB) axis labels on a white background. Under WCAG AA, what contrast ratio should normal-size text reach?
Answer: A. At least 4.5:1
WCAG AA asks for 4.5:1 for normal text and 3:1 for large text and key graphical parts, such as chart lines. #BBBBBB on white is about 1.9:1, too faint for many readers. Use a darker grey and check with a contrast tool.
Sales staff in the field open a 12-visual desktop dashboard on their phones, and everything is tiny. What is the best approach?
Answer: D. Build a phone layout with a few key visuals stacked
Phones need a separate layout: the 3–5 things field staff check most, stacked vertically, with large touch targets and short labels. Power BI and Tableau both support device-specific layouts for this.
Order values from two customer types (small shops and wholesalers) form two peaks. With very wide bins, the histogram shows one hump. What does this show about histograms?
Answer: A. Bin width can hide or invent shape; try several
Bins that are too wide merge separate peaks; bins too narrow make noise look like structure. Try a few widths, or overlay a density curve, and split by group when you suspect different populations.
A Tableau workbook is published for 200 sales reps, and each should see only their own accounts. What is a common, scalable approach?
Answer: C. An entitlement table filtered on USERNAME()
An entitlement table maps each username to the accounts they may see. A data source filter like [Rep Username] = USERNAME() then restricts every view. Adding a rep means adding rows, not editing the workbook.
A dashboard shows order status using only green and red dots. What is the best accessible fix?
Answer: B. Add text labels or distinct icons so colour is not the only way to understand status
WCAG says colour must not be the only way to convey information. About 1 in 12 men has some colour vision deficiency, and screen readers cannot see colour. Text or icons, with colour as extra support, work for everyone.
A bar chart uses thick black gridlines and a heavy border, and the bars are hard to read. What is the usual fix?
Answer: B. Lighten or remove the gridlines
Non-data elements should support, not compete with, the data. Light grey or fewer gridlines and no heavy frame keep reference marks available while the bars become the clear figure.
The Pearson correlation between discount level and customer satisfaction is 0.02. A colleague says discounts have no relationship with satisfaction. What should you check?
Answer: D. Plot the data: Pearson only measures straight-line relationships, so a U-shape or a pattern hidden inside segments can still give a value near zero
Pearson correlation captures linear association only. If satisfaction rises with small discounts but falls with very large ones (perhaps because they signal problems), the effects cancel and r is near zero. A scatter plot, a segment breakdown or a rank-based measure reveals the pattern.
Research by Cleveland and McGill ranked how accurately people compare values. Which encoding is judged most accurately?
Answer: A. Position along a common scale
People compare positions on a shared axis, as in bar and dot charts, far more accurately than angles, areas or colour intensity. That is why bars usually beat pies and bubbles when exact comparison matters, and colour is best for grouping or highlighting.
Store 1 had 50 orders from 100 visits; store 2 had 10 orders from 1,000 visits. A card shows "Avg conversion 25.5%". What is the overall conversion rate?
Answer: D. 5.45%
Averaging the two rates (50% and 1%) gives 25.5%, treating a tiny store and a big one equally. The overall rate is total orders over total visits: 60 / 1,100 ≈ 5.45%. Compute ratios as SUM(numerator) / SUM(denominator) at each level.
A slide shows marketing spend (left axis ₹0–10 Cr) and sign-ups (right axis 40,000–42,000) rising together and claims spend drives sign-ups. What is the issue?
Answer: A. The right axis is cropped to exaggerate a tiny change
Sign-ups barely move (about 5%), but a 40,000–42,000 axis stretches that small rise to match the spend line. Matching shapes on a dual axis prove nothing. Show both as % change from a base, and test the relationship properly.
A chart claims "Sales up 40%" by starting in April 2020, during lockdown. Over 5 years sales are roughly flat. What makes such a chart honest?
Answer: B. Show a longer window; flag the odd base
Picking a low starting point makes almost any series look like growth. Show enough history to judge the trend, or compare with a normal base period, and call out unusual periods such as lockdowns so readers understand the context.
A bar chart shows average order value for two app versions with 95% confidence interval error bars that overlap slightly. A manager says "so there is no significant difference". Is that right?
Answer: D. Not always; slight overlap can be significant
Two 95% intervals can overlap a little while the difference between the means is still significant at 5%; the test uses the standard error of the difference, which is smaller than the sum of both margins. Show the interval for the difference itself if that is the question.
A pie of three channels shows labels 34%, 33% and 34%, which add up to 101%. A board member asks if the data is wrong. What is the likely explanation?
Answer: B. Each share was rounded on its own
Values like 33.6%, 32.8% and 33.6% each round to 34%, 33% and 34%, which add up to 101%. Show one decimal place, use a largest-remainder method so labels add to 100%, or add a note that figures are rounded.
A stacked area chart shows monthly sales of 4 channels. The top layer's band looks like it grows sharply. Why can that be misleading?
Answer: C. Each band's shape follows the layers below it
Only the bottom layer sits on a flat baseline. Upper bands ride on the ups and downs of the layers beneath, so their edges move even when their own value is flat. Judging their trend means judging changing thickness, which is hard. Use lines or small multiples for each channel's trend.
A dashboard shows a trailing 28-day average of daily sales. Sales dropped sharply on 1 March and stayed low. Why does the chart only show half of the drop around 15 March?
Answer: D. Each point averages 28 past days
On 15 March the window holds 14 days before the drop and 14 after, so the average shows only half the fall, and the full drop appears after 28 days. Longer windows smooth more but react later; show the raw series or a shorter window alongside for recent changes.
A choropleth of election turnout makes large, sparsely populated regions dominate the map, while dense city areas are tiny and hard to see. What is a good alternative?
Answer: B. A tile map or a population cartogram
On a geographic map, visual weight follows land area, not people. Tile grid maps give each region the same size, and cartograms size regions by population, so small dense areas get fair attention. A bar chart sorted by value is another honest option.
An employee survey has 12 questions answered on a 5-point scale from Strongly disagree to Strongly agree. Leaders want to see which questions lean positive or negative. Which chart suits this best?
Answer: A. A diverging stacked bar centred on the neutral answer
A diverging stacked bar puts disagree shares to the left of a centre line and agree shares to the right, so the lean of each question is visible and questions can be sorted by it. Twelve pies are hard to compare, and an average hides how split the answers are.
In a Tableau view blended on Region, a secondary field Manager shows "*" for some regions. What does the asterisk mean?
Answer: C. ATTR() found several values at that level
Secondary fields are aggregated, and dimensions use ATTR(), which returns a value only if it is the same for every row. When a region has two managers, ATTR returns *. Link on more fields or change the level of detail.
A 4 billion row fact table must stay in DirectQuery, but 90% of report queries are by month, region and product category. What makes those fast?
Answer: B. An imported aggregation table at that grain
User-defined aggregations let an imported summary table answer queries at its grain, while queries needing detail fall through to the DirectQuery fact. Users see one model; most clicks hit the fast in-memory summary.
In Microsoft Fabric, a semantic model uses Direct Lake on Delta tables in OneLake. How does it differ from Import mode?
Answer: B. It pages Delta files into memory as needed
Direct Lake reads the Parquet files of Delta tables straight into the in-memory engine when queried, so there is no scheduled import copy, yet queries run at near Import speed. If some limits are exceeded it can fall back to DirectQuery.
A Gulf retailer's month-on-month sales change chart swings wildly every year around Ramadan and Eid. Leaders want to know if the business is growing. Which view helps most?
Answer: B. Year-on-year change for each month or festival
Seasonal peaks make month-on-month changes reflect the calendar, not growth. Comparing each period with the same period last year cancels most of the seasonality. Because Ramadan moves about 11 days earlier each year, align by the festival period too.
A heatmap of sales by store and week uses a linear colour scale. One store had a ₹50 crore bulk order in one week, and every other cell now looks the same pale colour. What helps?
Answer: D. Cap the scale or use quantile or log bins
With a linear scale the outlier uses up the whole colour range. Capping at a high percentile, using quantile bins, or a log scale restores contrast among normal cells; mark the outlier separately so it is not hidden.
A call centre dashboard ranks agents by average handle time. Within a month, handle time falls 30% but repeat calls and complaints rise. What is the best dashboard change?
Answer: D. Pair it with balancing metrics like resolution
When one number becomes a target, people optimise it, here by rushing calls (Goodhart's law). Showing counter metrics such as first-call resolution and satisfaction next to it keeps the trade-off visible and discourages gaming.
A line chart compares 12 emirate-level sales series, each in a different hue, and nobody can match lines to the legend. What usually works better?
Answer: B. Highlight a few key series and grey the rest
People can reliably tell apart only a handful of categorical colours. Emphasising the series that matter, labelling lines directly and muting the rest, or splitting into small multiples, keeps the chart readable.
You are designing an infographic about health schemes for a government website. Many citizens use screen readers or have low vision. What is essential?
Answer: D. Ensure good contrast and readable text sizes, do not rely on colour alone, and provide the same information as accessible text or HTML
Text inside an image cannot be read by screen readers and does not resize well. Public-sector content must meet accessibility standards, so provide a text alternative, use strong contrast, readable sizes and labels or icons alongside colour.
An infographic needs to compare market share for 12 companies, several with similar values. The client asks for a pie chart. What is the better option?
Answer: C. A sorted horizontal bar chart with direct labels, possibly grouping the smallest into "Other"
People compare lengths more accurately than angles or areas, and many similar slices become impossible to read. Sorted bars with labels make ranking and differences clear and avoid relying on colour legends.
Take a 15-question Data Visualisation MCQ test with a score at the end, or practise the full Data Visualisation question bank and earn points on the leaderboard.
