Not sure which lesson to do first? Find your level and get a plan →

Math · Statistics & Probability · Grade 6-8

Misleading Graphs

A misleading graph distorts the true pattern through tricks like a truncated axis, unequal intervals, cherry-picked data, or a stretched scale — the numbers may be real but the picture lies.


Quick Answer

A misleading graph distorts the true pattern through tricks like a truncated axis, unequal intervals, cherry-picked data, or a stretched scale — the numbers may be real but the picture lies. Use this lens whenever a chart is trying to persuade you: an ad, a news graphic, a campaign flyer. Read the axis numbers before the shapes, check that a bar graph's axis starts at zero, look for uneven steps, and ask what time period or data was left out. Then compare the real numbers with the impression the picture gave.

Why This Matters

Graphs persuade in seconds, before anyone reads the numbers. A sales chart that starts its axis at 95 can make a 2% rise look like a doubling, and a chart showing only the good months can hide a bad year. Knowing the common tricks lets you read news and ads critically and draw honest graphs yourself.

Intuitive Explanation

Two brands of cereal sell 102 and 108 thousand boxes. Drawn from 0, the bars look almost the same height, which is the truth: 108 is about 6% more than 102. Now start the vertical axis at 100. The first bar is 2 units tall and the second is 8, so the second bar looks four times as tall. The numbers did not change; only the picture did.

Most misleading graphs use one of a few tricks. A truncated axis starts a bar graph above zero, so small differences look huge. Unequal intervals space the axis labels unevenly, such as 0, 10, 20, 50, 100, which bends a trend. Cherry-picking shows only the time period or groups that support a story, such as the 4 good months of a bad year. Picture graphs can mislead too: doubling the height of a drawn money bag also doubles its width, so it covers four times the area.

Not every surprising graph is misleading. A line graph of temperature can sensibly start at 10 °C, because zero has no special meaning there, and a graph with a real outlier looks dramatic because the data is dramatic. To call a graph misleading, name the specific trick and show the gap between what the picture suggests and what the numbers say.

Think of it like this. A misleading graph works like a fun-house mirror. The reflection really is you, every feature is there, but the curved glass stretches some parts and squeezes others. A misleading graph keeps the real numbers and bends how they look. To see yourself accurately you check a flat mirror; to see data accurately you check the axis numbers.

Core idea. A misleading graph uses real data but distorts the design so you read the wrong story.


When to Use

Look for misleading design when a graph is meant to persuade, when a difference looks dramatic, or when a bar graph's axis does not start at zero. Also check for uneven intervals, missing time periods and drawn pictures that grow in two directions at once. If the only problem is the axis baseline or spacing, call it scale distortion specifically; if the graph is honest but the data is surprising, it may simply be a real result.

Tip. Ask: Does the picture exaggerate or hide a difference that the actual numbers don't support?

How to Recognize It

Read the numbers on the axes and check which dates or groups are shown before you look at the shapes, then compare the story the picture tells with the story the numbers tell.

  1. Do the drawn sizes match the ratio of the real values?

    Measure them: if one bar is 3 times as tall but its value is only 10% bigger, the design is distorting. When the trouble is only where the axis starts or how its steps are spaced, call it scale distortion.

  2. Does the graph show only part of the time period, or only some of the groups?

    Yes means check what was left out: three months of rising sales picked from a year that fell overall is cherry-picking, misleading even with a perfect axis. If everything is shown, the trend may be real.

  3. Are pictures or 3-D shapes used instead of plain bars?

    Yes means look at area, not just height: a coin drawn twice as tall and twice as wide covers 4 times the area, so it looks 4 times bigger for a value only twice as big.

  4. Is the surprising part just one value that sits far from the rest?

    Yes means it may be an outlier, a real data point, not a design trick. A graph that shows an outlier honestly is not misleading; one that hides it, or cuts it off the axis, can be.

Misleading Graphs compared with Scale distortion, Data visualization (honest), and Outlier

All four rows can make a chart look dramatic. What separates them is where the drama comes from: a design choice in general (misleading graph), the axis in particular (scale distortion), one real extreme value (outlier), or the data themselves drawn fairly (honest visualization).

Misleading Graphs

Meaning
Use this when you must judge whether a chart's design honestly represents the data behind it.
Key test
Does the picture exaggerate or hide a difference that the actual numbers don't support?
Example
A graph shows company profit bars: Year 1 at $102M looks far shorter than Year 2 at $108M — about a quarter its height. The axis starts at $100M. Is the picture fair?

Scale distortion

Meaning
The specific axis/interval trick, one type of misleading graph.
Key test
Use when the distortion is precisely the axis not starting at zero or uneven intervals.
Example
Y-axis runs 90–100 instead of 0–100

Data visualization (honest)

Meaning
An accurate graph designed to reveal, not distort, the pattern.
Key test
Use when making or reading a fair chart, not critiquing a rigged one.
Example
Full-axis bar graph of survey results

Outlier

Meaning
A genuine extreme data point, not a design trick.
Key test
Use when one real value is far from the rest, not when the graph is rigged.
Example
One $1,000 sale among $50 sales

Worked Examples

Example 1 — Catch the trick

Easy

A graph shows company profit bars: Year 1 at $102M looks far shorter than Year 2 at $108M — about a quarter its height. The axis starts at $100M. Is the picture fair?
Show the steps and the answer
  1. Read the values and the axis: Year 1 is $102M, Year 2 is $108M, and the axis starts at $100M.

    Bar heights compare honestly only when they are measured from zero.

  2. Find the drawn heights: 102−100=2 and 108−100=8 units, so Year 2's bar is 8÷2=4 times as tall.

    With the axis cut at 100, each bar shows only the part above $100M.

  3. Find the real change: 108−102=6, and 6÷102≈0.06, about 6%.

    The actual growth is a small rise, not the fourfold jump the bars suggest.

  4. Compare the picture with the numbers: a bar 4 times as tall for a 6% increase.

    When the drawn ratio and the real ratio disagree this much, the graph is misleading.

Answer: Misleading — the truncated axis exaggerates a 6% rise

Takeaway: On a bar graph, check where the axis starts. A cut-off baseline here turns a 6% rise into a picture of a fourfold jump.

Example 2 — Honest version

Standard

The same profits graphed with the axis starting at $0. Is this graph misleading?
Show the steps and the answer
  1. Find the drawn heights from $0: 102 and 108 units.

    With a zero baseline, each bar's height is the full value.

  2. Compare the ratio: 108÷102≈1.06, so Year 2's bar is about 6% taller.

    The visual difference now matches the real difference.

  3. Check the other tricks too: the same two years are shown, the axis steps are even, and the bars are plain and flat.

    A zero baseline is not the whole test; a graph is honest only if no other distortion is present.

Answer: Not misleading — it's an honest visualization

Takeaway: A bar graph drawn from zero with even steps shows differences at their true size, even when the difference is small.

Example 3 — Find the mistake: real numbers, false story

Application

A city graph titled 'Ridership soaring' shows monthly bus riders rising from 40,000 in March to 52,000 in June. The full data shows 60,000 riders in January, falling to 40,000 in March before rising to 52,000 in June. Tom writes: "Every number on the graph is real, so it can't be misleading: ridership is soaring." Find the mistake and correct it.
Show the steps and the answer
  1. Agree with what is true: 40,000 in March and 52,000 in June are real figures.

    Real numbers are needed for an honest graph, but the choice of which numbers to show matters too.

  2. Look at the full period: January 60,000 and June 52,000, a change of 52000−60000=−8000.

    From January to June, ridership actually fell.

  3. Size the shown rise against the hidden fall: up 52000−40000=12000 since March, after a drop of 60000−40000=20000.

    The graph starts at the lowest month, so a partial recovery looks like growth.

  4. Correct the graph: show January to June, with a title such as 'Ridership recovering, still below January'.

    Showing the whole period lets readers see both the drop and the recovery.

Answer: The graph misleads by cherry-picking: from January to June ridership fell from 60,000 to 52,000, so 'soaring' is false even though every number shown is real.

Takeaway: Real numbers can still tell a false story. Check what time period or data was left out.

Where You'll See It

  • A TV chart about tax rates

    In 2012 a US news channel showed a bar chart comparing a top tax rate of 35% with a planned 39.6%, with the axis starting at 34%. Above the baseline the bars stood 1 and 5.6 units tall, so the second looked more than 5 times as tall, although the rate was rising by 4.6 percentage points, about 13%.

  • Picture graphs in advertisements

    An ad draws a small milk carton for last year's sales and one twice as tall for this year's. Because the drawing doubled in both height and width, it covers four times the area, so readers sense a much bigger jump than the real doubling.

  • Choosing a convenient start year

    Some graphs in debates about global warming started in 1998, an unusually hot El Niño year, so the next decade looked flat. Graphs covering 1970 onward show the steady long-term rise, and later years were warmer still.

Common Mistakes

  1. The slip: Trusting a graph because the data is real

    Instead: the distortion lives in the design choices, not the numbers.

  2. The slip: Only checking the bars and ignoring the axis

    Instead: the truncated axis is the most common trick.

  3. The slip: Assuming any surprising graph is misleading

    Instead: confirm a specific distortion (axis, interval, scale) before crying foul.


Mini Practice

Try these on your own. Tap Reveal when you want to check.

  1. A bar graph shows Brand A with 48 votes and Brand B with 52, but the axis starts at 46. How many times taller does B's bar look? How much bigger is B really?

    Hint: Measure each bar from 46, then compare the real numbers.

  2. A picture graph shows a small coin for 100 sales and a coin twice as wide and twice as tall for 200 sales. How many times bigger does the second coin look?

    Hint: Think about area, not just width.

  3. Monthly sales were Jan 50, Feb 45, Mar 40, Apr 42, May 44. An ad graphs only March to May with the title "Sales rising." Is that fair?

    Hint: Compare January with May.

  4. A line graph's time axis shows 2000, 2010, 2015, 2018 and 2020 evenly spaced. Why does recent growth look faster than it is?

    Hint: How many years does each equal step cover?

  5. A bar graph of two schools' test averages, 71 and 74, starts at 0, has a labeled scale and shows two nearly equal bars. Is it misleading?

    Hint: Look for a specific trick: a cut axis, uneven intervals or missing data.

  6. Profits were 20, 21 and 22 million dollars, but the axis runs from 19.5 to 22. Lee says: "The numbers are real, so the graph cannot mislead." Correct him.

    Hint: Compare the bar heights above 19.5 with the real growth.

Want the full set?

50 practice questions for this concept — free to try, every one with a complete worked solution showing the why, not just the answer.

Frequently Asked Questions

Is it misleading to leave out data that is not relevant?

Not if the choice is fair and stated. A graph about this season's football scores can leave out last season. It becomes misleading when the cut changes the story: a company showing only the three months its sales rose, from a year when they fell, is choosing the window to get the conclusion it wants. Ask whether adding the missing data would change what a reader believes.

Why do pie charts sometimes mislead even when the slices are correct?

There are two common ways. A tilted 3-D pie makes the front slices look larger than slices of the same size at the back. And a pie only makes sense when the slices are parts of one whole that add to 100%; a poll where people could pick several answers can add to 130%, and squeezing that into a pie gives every slice the wrong size.

Can a graph mislead by using two different vertical axes?

Yes. When two lines share one picture but each has its own scale, one on the left and one on the right, the maker can stretch either scale until the lines cross, run together, or seem to follow each other. The crossing point means nothing, because it depends on how the scales were chosen. Compare the actual numbers, or redraw both lines as percent changes from the same start.

Is a misleading graph always made on purpose?

No. Some graphing programs pick an axis that starts just below the smallest value when the numbers are close together, so a careless maker can produce a distorted bar graph without meaning to. The effect on the reader is the same either way, which is why you check the axes yourself instead of trying to guess what the maker intended.

Learning Path

← Before

Data Visualization
Misleading Graphs

You are here

You need Data Visualization first: how bar graphs, line graphs and pictographs show data. Misleading Graphs adds a reader's checklist for distorted designs, from truncated axes to cherry-picked data. Next, Scale Distortion looks closely at the most common trick, the axis baseline and spacing, and shows how to measure exactly how much it exaggerates.

See Also