Tableau Quiz: Histograms And Box Plots
10 questions · exam conditions
0:00
Histograms And Box PlotsQuestion 1 of 10

A worksheet displays individual customer sales marks, with Region dividing the view into separate panes. The analyst wants Tableau to calculate a separate box-and-whisker distribution for each region rather than one distribution across the entire worksheet.

When dragging Box Plot from the Analytics pane, which drop target should the analyst use?

Use Table so all regional marks are pooled into a single worksheet-level box plot.
Use Pane so Tableau computes an independent box plot within each regional pane.
Use Cell so Tableau computes the distribution at a finer partition than the pane level.
Use Table and then manually partition results by adding Region to the box plot's scope settings.
← Back to quizzes

Tableau Quiz

Tableau Quiz: Histograms And Box Plots

Practice Histograms And Box Plots in Tableau with focused quiz questions that help you check what you know, review explanations, and build confidence with test-style prompts.

What this quiz covers

This quiz focuses on Histograms And Box Plots, giving you a quick way to practice the rules, question types, and explanations that matter most for Tableau.

How to use this quiz

Try each quiz question before looking at the correct answer. Use the explanations to review missed ideas, then come back to similar questions until the pattern feels familiar.

All questions

Question 1

A worksheet displays individual customer sales marks, with Region dividing the view into separate panes. The analyst wants Tableau to calculate a separate box-and-whisker distribution for each region rather than one distribution across the entire worksheet.

When dragging Box Plot from the Analytics pane, which drop target should the analyst use?

  1. Use Table so all regional marks are pooled into a single worksheet-level box plot.
  2. Use Pane so Tableau computes an independent box plot within each regional pane. (correct answer)
  3. Use Cell so Tableau computes the distribution at a finer partition than the pane level.
  4. Use Table and then manually partition results by adding Region to the box plot's scope settings.
Explanation: When dragging a reference object like a Box Plot from the Analytics pane, Tableau asks you where to scope the calculation — meaning, which partition of data should define each independent distribution. The three drop targets (Table, Pane, Cell) correspond to progressively finer levels of granularity in your view. Because the analyst wants a separate box-and-whisker plot for each region, and Region is already dividing the view into distinct panes, dropping onto Pane is exactly right. Tableau will treat each regional pane as its own dataset and compute an independent five-number summary (minimum, Q1, median, Q3, maximum) within that boundary — which is precisely what B describes. A is wrong because dropping onto Table pools all marks across the entire worksheet into one box plot, completely ignoring the regional pane structure. You'd get a single distribution that blends all regions together, which defeats the analyst's goal. C is wrong because Cell operates at a finer grain than a pane — it would subdivide calculations within individual intersections of row and column headers, which is more granular than needed here and would fragment the distribution further than intended. D is a fabricated option; Tableau's drag-and-drop workflow for Analytics pane objects doesn't include a "scope settings" configuration step after placing at the Table level. That's not how the feature works. A good rule of thumb: mentally match the drop target to the visual boundary you want each calculation to respect. If your goal lives at the pane level, drop at Pane — the target names are quite literal in Tableau's Analytics pane workflow.

Question 2

An analyst has a histogram of Sales using a common Sales (bin) field. The analyst must compare the shapes of the Consumer, Corporate, and Home Office distributions in separate aligned sections while preserving identical bin boundaries.

Which modification best supports the comparison?

  1. Place Segment on Filters and select all segments so Tableau merges them into one distribution.
  2. Place Segment on Color and enable stacking so all segment frequencies accumulate in each shared bar.
  3. Create a different bin field for each segment and allow Tableau to choose separate bin sizes.
  4. Place Segment on Rows while retaining Sales (bin) on Columns and the frequency measure on Rows. (correct answer)
Explanation: When comparing distributions across categories in Tableau, your goal is to create small multiples — separate, aligned charts that share the same axes so shapes are directly comparable. This is the core concept being tested here. Placing Segment on Rows while keeping Sales (bin) on Columns and frequency on Rows (answer D) does exactly this. Tableau generates three horizontally aligned histograms — one per segment — stacked vertically. Because all three panels share the same bin field and the same x-axis scale, the bin boundaries are identical across segments, making shape comparison straightforward and honest. Answer A fails because filtering to "all segments" doesn't separate them — it simply shows the combined distribution as a single merged histogram, which masks the individual shapes you're trying to compare. Answer B stacks segment frequencies within each shared bar using Color. While this does overlay segments, stacked bars make it very difficult to compare the shapes of individual distributions because each segment's bars start at different baselines. You can't easily judge whether Corporate's distribution is skewed the same way Consumer's is. Answer C is the most dangerous distractor. Creating separate bin fields per segment allows Tableau to calculate different bin sizes for each group. This destroys comparability — if Consumer uses $100 bins and Corporate uses $250 bins, the shapes are no longer on the same scale and any visual comparison is misleading. Study tip: Whenever a Tableau question asks about comparing distributions across categories, think "small multiples" first — a dimension on Rows or Columns creates aligned panels that share axes, which is almost always the cleanest solution.

Question 3

An analyst creates a histogram from a numeric bin field. The data contains observations in bins labeled 00, 1010, and 3030, but no observations in the bin labeled 2020. The analyst wants the empty interval to remain visibly represented so viewers can distinguish an empty bin from a compressed axis.

Which action best meets the requirement without creating artificial source records?

  1. Turn on Show Missing Values for the binned field so the missing interval is included in the domain and renders as a zero-height bar. (correct answer)
  2. Click the null indicator at the bottom of the view so Tableau converts the empty interval into a separate null-valued bar in the histogram.
  3. Select Analysis > Stack Marks > Off so the empty interval receives its own unstacked position alongside the other bins.
  4. Change the frequency measure from COUNT to COUNTD so Tableau generates a placeholder record for the absent bin.
Explanation: When working with binned fields in Tableau histograms, the core challenge is controlling which domain values appear in the view — including intervals where no data exists. Tableau's Show Missing Values option is specifically designed for this: when you right-click a discrete or binned dimension on the axis and enable it, Tableau forces every interval in the bin domain into the view, rendering empty bins as zero-height bars rather than silently skipping them. This is exactly what the scenario requires — a visible gap at bin 2020 without touching the underlying data source. Choice A is correct because Show Missing Values operates at the field-display level, expanding the axis domain to include absent intervals while leaving source records completely unchanged. Choice B describes a real UI element — the null indicator — but clicking it filters or includes null values, not structured empty bins. A missing bin interval is not the same as a null record; conflating the two is a classic trap on this exam. Choice C, toggling stack marks off, controls how multiple measures overlap visually. It has no effect on which dimension members appear along the axis, so it cannot make a missing bin visible. Choice D is also wrong because switching COUNT to COUNTD (distinct count) only changes how existing records are aggregated. Neither function generates placeholder records for bins that have zero observations — the empty interval still disappears from the view. As a study tip: whenever a Tableau question mentions missing axis values or empty bins, your first instinct should be Show Missing Values on the dimension — it's the purpose-built solution for domain completeness without data manipulation.

Question 4

A worksheet contains thousands of marks behind a box-and-whisker plot. The analyst wants a cleaner presentation that shows the boxes, whiskers, and outlier marks but suppresses the dense cloud of non-outlier marks.

Which box plot option best satisfies this requirement while preserving the distribution calculation?

  1. Set the mark opacity to zero so the box plot and all outliers become transparent.
  2. Filter all marks between the lower and upper whiskers out of the data source.
  3. Disable Show Mark Labels so non-outlier marks are removed from the worksheet.
  4. Enable Hide underlying marks (except outliers) in the box plot options. (correct answer)
Explanation: When working with box-and-whisker plots in Tableau, you need to distinguish between display options and data manipulation. The goal here is purely presentational — hiding cluttered marks without changing what the box plot actually calculates. Tableau's box plot formatting includes a built-in toggle called Hide underlying marks (except outliers), which does exactly what the analyst needs: it suppresses the dense cloud of individual marks that fall within the whisker range while keeping outlier marks visible. Crucially, the underlying data still drives the distribution calculation — the boxes and whiskers remain accurate. Option D is the correct choice because it's a targeted, non-destructive display setting designed precisely for this use case. Option A fails because setting mark opacity to zero hides everything, including the outlier marks the analyst specifically wants to preserve. You'd lose the very data points that make outlier analysis meaningful. Option B is a deeper mistake — filtering marks out of the data source between the whisker boundaries would corrupt the distribution calculation itself. Tableau computes quartiles and whiskers from the full dataset, so removing values would shift the box positions and whisker lengths, misrepresenting the data entirely. Option C is a trap based on confusing terminology: Show Mark Labels controls whether text labels appear on marks — it has no effect on whether the marks themselves are rendered or hidden. A useful study pattern: on the Tableau exam, questions about box plots often test whether you know which settings are display-only versus which ones alter data or calculations. Always ask yourself — does this change what Tableau computes, or just what it shows?

Question 5

A data source contains one row per completed delivery and a numeric measure named Delivery Days. An analyst wants a histogram in which each bar represents a range of delivery times and the bar height represents the number of deliveries in that range.

Which Tableau configuration most directly produces the requested histogram?

  1. Place Delivery Days (bin) on Columns and CNT(Delivery Days) on Rows. (correct answer)
  2. Place continuous Delivery Days on Columns and SUM(Delivery Days) on Rows.
  3. Place Delivery Days on Detail and AVG(Delivery Days) on Rows.
  4. Place Delivery Days (bin) on Columns and AVG(Delivery Days) on Rows.
Explanation: When building a histogram in Tableau, you need two things working together: a way to group continuous values into ranges (bins), and a count of records within each range. Keeping that in mind makes this question straightforward. A histogram requires bins on the x-axis and a frequency count on the y-axis. Tableau's bin field — created by right-clicking a measure and selecting "Create Bins" — groups individual values into equal-width intervals, which is exactly what each bar in a histogram represents. Placing Delivery Days (bin) on Columns creates those intervals, and CNT(Delivery Days) on Rows counts how many deliveries fall into each one. That's precisely the bar height the analyst wants, making A the correct answer. Choice B places the raw, continuous Delivery Days field on Columns with SUM(Delivery Days) on Rows. Without bins, Tableau treats each unique value as its own mark rather than grouping them into ranges, and summing delivery days gives you a total time — not a count of deliveries. Choice C puts Delivery Days on Detail, which only adds granularity to existing marks; it doesn't create bars over ranges, and AVG still measures duration rather than frequency. Choice D is the closest distractor — it correctly uses the bin field on Columns — but AVG(Delivery Days) on Rows returns the average delivery time within each bin, not the count of deliveries. The bar heights would represent duration, not frequency, so it's not a histogram. A helpful rule of thumb: any true frequency histogram pairs a bin field on one axis with a COUNT on the other. If either piece is missing or substituted, it's no longer a histogram.

Question 6

A worksheet has Region on Columns and SUM(Sales) on Rows. Because the view contains only one aggregated mark per region, adding a box plot does not provide a meaningful distribution. Each order should be one observation in the distribution.

What should the analyst do before adding the box plot?

  1. Add Order ID to Detail so each region contains order-level marks for the box plot. (correct answer)
  2. Add Region to Detail so Tableau duplicates each regional aggregate into multiple observations.
  3. Convert SUM(Sales) to AVG(Sales) so quartiles can be computed from the single mark.
  4. Hide the Region header so Tableau combines the regional aggregates into one distribution.
Explanation: When building a box plot in Tableau, the key concept to understand is granularity — how many marks exist per partition. A box plot needs multiple observations per category to compute quartiles, whiskers, and outliers. If your view only has one aggregated mark per region (one SUM(Sales) bar), Tableau has nothing to distribute; there's no spread to visualize. The fix is answer A: adding Order ID to the Detail shelf. This disaggregates the data so that each individual order becomes its own mark within its region. Now Tableau has a collection of order-level sales values per region, giving the box plot the distribution it needs to calculate meaningful statistics like the median and interquartile range. B is wrong because adding Region to Detail doesn't create new observations — it just references the same field already driving your columns. Tableau won't manufacture extra marks by duplicating a dimension that's already in play. C is a misconception about aggregation: switching from SUM to AVG still produces a single mark per region, just with a different value. One number cannot form a distribution regardless of how it's calculated. D is a red herring — hiding the Region header doesn't merge or redistribute data; it only affects visual labeling, leaving the underlying mark structure unchanged. Study tip: Whenever a question mentions adding a statistical chart type (box plot, histogram), immediately ask yourself, "How many marks exist per partition?" If the answer is one, you need to increase granularity by adding a dimension to Detail — that's almost always the correct move.

Question 7

A sales data source contains one row per line item, and an order may contain several line items. The requested histogram must show the distribution of total sales per order, with each order counted exactly once.

Which approach produces the required histogram at the correct grain?

  1. Bin row-level Sales and use COUNT(Sales) as the frequency, counting each line-item row as a separate observation in the interval.
  2. Bin the fixed order total and use COUNT(Order ID) as the frequency so that orders with multiple line items each contribute multiple observations to their bin.
  3. Calculate order total with a fixed Order ID expression, bin that result to define order-total intervals, and use COUNTD(Order ID) as the frequency so each order is counted once. (correct answer)
  4. Place SUM(Sales) directly on Columns and COUNTD(Order ID) on Rows without binning, relying on the continuous axis to represent the distribution.
Explanation: When building a histogram in Tableau from transaction-level data, the core challenge is granularity: you need one value per order before you bin, otherwise row-level duplicates corrupt the distribution. The right approach has three steps: collapse each order to a single total, bin those totals into intervals, then count distinct orders per bin. Option C does exactly this — a FIXED LOD expression like {FIXED [Order ID] : SUM(Sales)} computes one total per order regardless of how many line items exist. Binning that calculated field creates meaningful intervals, and COUNTD(Order ID) ensures each order contributes exactly one count to its bin, giving you the true distribution of order totals. Option A fails at the first step: binning row-level Sales means a single $500 order split across five $100 line items would place five observations into the $100 bin instead of one observation in the $500 bin — completely misrepresenting the distribution. Option B recognizes the need to compute an order total, but then contradicts itself by using COUNT(Order ID) instead of COUNTD(Order ID). For a multi-line-item order, COUNT still returns multiple rows, so that order gets counted more than once in its bin — the same duplication problem as A. Option D skips binning entirely. Placing SUM(Sales) on Columns produces a single aggregate bar, not a histogram with intervals. Without bins, there is no frequency distribution to display. Study tip: On histogram questions, ask yourself three things in order — What is one observation?, How do I aggregate to that grain?, and How do I count without duplication? FIXED LOD + COUNTD is the classic Tableau answer when "one observation" is an order, not a row.

Question 8

A box plot displays customer-level profit marks by segment. A standard dimension filter is then used to remove one customer type containing several high-profit observations. No fixed level-of-detail calculations or table calculations are used.

What should the analyst expect after applying the filter?

  1. The marks disappear, but the original quartiles and whiskers remain fixed until the workbook is reopened.
  2. The box plot recalculates its quartiles and whiskers from the observations remaining after the filter. (correct answer)
  3. Only the whiskers recalculate; the median and quartiles remain based on the unfiltered observations.
  4. The box plot becomes invalid because analytic objects cannot respond to dimension filters.
Explanation: When working with box plots and filters in Tableau, the key concept to keep in mind is how dimension filters interact with the underlying data that drives analytic objects like box plots. Ask yourself: does the filter change the dataset the visualization is built from, or does it only affect visual display? Dimension filters in Tableau operate at the data level — they remove records from consideration before any aggregation or statistical computation occurs. A box plot's quartiles, median, and whiskers are all derived from the actual observations in the view. When a dimension filter removes a customer segment, those marks are gone from the dataset powering the chart, so Tableau recomputes the entire box plot — Q1, Q3, median, and whiskers — using only the remaining observations. That's exactly what answer B describes, making it correct. Answer A is wrong because Tableau does not "freeze" statistical calculations until a workbook is reopened. There is no such caching behavior for box plots — they are dynamic and recalculate with each data change. Answer C is incorrect because it implies a split behavior where only whiskers update while the median and quartiles stay fixed. Tableau makes no such distinction; all components of the box plot recalculate together from the same filtered dataset. Answer D is a fabricated rule — analytic objects like box plots and reference lines absolutely respond to dimension filters, and there is no concept of a box plot becoming "invalid" due to filtering. As a study tip, remember that in Tableau, standard dimension filters sit near the top of the order of operations, meaning they shape the data before nearly everything else — including statistical summaries — so what you filter out simply no longer exists for the viz to calculate from.

Question 9

A Tableau box plot is configured to use whiskers extending to 1.51.5 times the interquartile range. For one category, the first quartile is 1010 and the third quartile is 1818. The category contains observations at −4-4, 00, 1010, 1818, 2929, and 3131.

Which statement correctly describes the upper whisker and upper outlier classification?

  1. The upper whisker ends at 3030, and 3131 is plotted as the only upper outlier.
  2. The upper whisker ends at 2929, and 3131 is plotted as an upper outlier. (correct answer)
  3. The upper whisker ends at 3131 because whiskers always extend to the observed maximum.
  4. The upper whisker ends at 1818 because the third quartile is the box plot's upper limit.
Explanation: When working with Tableau box plots, the whisker calculation follows a specific process: compute the IQR, multiply by 1.5 to find the fence, then locate the most extreme actual data point that still falls within that fence. The whisker ends at that real observation — not at the theoretical fence value. Here, the IQR is 18−10=818 - 10 = 8, so the upper fence is 18+(1.5×8)=18+12=3018 + (1.5 \times 8) = 18 + 12 = 30. Now scan the data above Q3 for the largest value that doesn't exceed 30: the candidates are 2929 and 3131. Since 31>3031 > 30, it falls outside the fence. The largest included value is 2929, so the upper whisker ends at 2929, and 3131 is plotted as an outlier. That's exactly what answer B describes. Answer A is tempting but wrong — 3030 is the fence, a calculated boundary, not a data point. Whiskers must snap to an actual observation in the dataset, and 3030 doesn't exist in the data. Answer C reflects a common misconception: whiskers do not automatically stretch to the maximum. They only reach the maximum if that maximum falls within the 1.5×IQR fence. Answer D confuses the box boundary with the whisker. The box's upper edge is Q3 (1818), but whiskers extend beyond the box to include non-outlier data points. A useful rule to memorize: fence first, then find the nearest real point inside it. On exam questions, always distinguish between the theoretical fence value and the actual data value where the whisker terminates.

Question 10

A histogram uses a bin size of 1010. Its visible bin headers, which represent lower limits, include 00, 1010, 2020, and 3030.

How will Tableau assign observations with values of exactly 29.929.9 and 3030?

  1. Both observations are assigned to the bin labeled 2020 because 3030 is its upper boundary.
  2. Both observations are assigned to the bin labeled 3030 because Tableau rounds values upward.
  3. The value 29.929.9 enters the bin labeled 2020, while 3030 enters the bin labeled 3030. (correct answer)
  4. The value 29.929.9 enters the bin labeled 3030, while 3030 enters the bin labeled 2020.
Explanation: When working with histograms in Tableau, the key concept to understand is how bin boundaries are handled — specifically, which bin "owns" the boundary value itself. Tableau uses left-closed, right-open intervals, meaning each bin captures values from its lower limit up to, but not including, the next bin's lower limit. So the bin labeled 2020 captures values where 20≤x<3020 \leq x < 30, and the bin labeled 3030 captures values where 30≤x<4030 \leq x < 40. This makes C the correct answer: 29.929.9 falls in [20,30)[20, 30) and lands in the 2020 bin, while 3030 falls exactly on the lower boundary of [30,40)[30, 40) and lands in the 3030 bin. Answer A is wrong because it assumes 3030 belongs to the 2020 bin as its upper boundary — but in Tableau's convention, the upper boundary is excluded, so 3030 does not belong to that bin. Answer B invents a rounding-up rule that Tableau does not apply; bin assignment is based on interval membership, not rounding. Answer D reverses the two assignments entirely, contradicting how left-closed intervals work — 29.929.9 is less than 3030 and stays in the lower bin, not the higher one. A useful memory aid: think of each bin label as a closed left door, open right door. Values walk in through the left (inclusive) and exit before reaching the right (exclusive). Whenever you see a question about exact boundary values in Tableau histograms, ask yourself which interval's left edge that value sits on.