Images and Charts as Argument
Photos and charts can support claims. Captions add context or suggest conclusions, scales shape comparisons, and crops leave details out. Check what the visual shows and whether it supports the claim made about it.
What a learner can do afterwards
- Say what a caption asks the reader to conclude that the image alone does not show
- Spot an axis or a scale that makes a difference look larger than it is
- Explain what a crop excludes and how the excluded part would change the reading
1 · Read
You have checked where written claims come from. Apply that habit to visuals. A caption may identify a place, date or event, or suggest a conclusion. Separate its words from what the picture shows. A photo of two skaters on one ramp cannot show how busy the whole park was. Check the source and wider context before repeating a wider claim.
Read axis labels and units before comparing sizes. On a linear scale, equal distances represent equal changes in value. The copied axis below claims to be linear, but its equally spaced ticks read 0, 5, 10 and 50. The changes are 5, 5 and 40 units. Its labels and spacing do not match. Other scales can be valid, but you must read their labels to understand the comparisons.
Bar lengths show value ratios when bars share units, a zero baseline and a linear scale. Twice the length then means twice the value. Draw values 50 and 55 from a baseline of 45, however, and their visible lengths become 5 and 10. One looks double, although 55 is not double 50. A nonzero axis is not always misleading: line graphs may use one to show small changes. Read the values before judging a change.
A crop removes part of an image. Compare it with the full photo: a close view of smiling spectators might leave out empty rows. Those rows weaken a claim about a full stadium, but do not erase the smiles. Missing details do not prove the opposite claim. A picture also does not settle a cause: a boarded-up cafe does not show why it closed or whether a town-wide policy worked.
Lines often show change over time; bars compare categories; pies show shares of a whole. Many pie slices can be hard to compare, but their number alone does not prove distortion. Check that slice sizes match the shares and labels identify them. In a picture chart, doubling an icon's height and width makes its area four times as large. If its value only doubled, that area can exaggerate the change.
Captions, axes, and crops each make claims you can test like sentences.
Lesson, questions and flashcards written for this dataset. Khan Academy visual-evidence articles and OpenStax Contemporary Mathematics, section 8.2, were studied for teaching structure and facts; their wording, figures and worked examples are not reproduced.
Four readings and the flashcards were written for this dataset. Video: Analyzing visual evidence (and vampires), from Khan Academy. The video's licensing terms have not been verified.
More to read
Separate the picture from the caption
Imagine a photograph of volunteers planting a tree beside a path. The caption calls it 'Residents restore the entire riverside.' You can see people planting a tree. Their identity and the condition of the whole riverside need evidence beyond this view.
Make two notes. In one, record the details visible in the photograph. In the other, record details or conclusions supplied by the caption. Compare the notes to find where the caption goes beyond the image.
Next, trace the photograph to its source. Check its date, location and any available information about changes to the published image. Look for records that address the wider restoration claim. Another page repeating the same caption is not an independent check.
A genuine image can support a narrow observation while leaving a broader claim unsettled. Report the supported observation and say which parts still need checking.
Read values before comparing bar lengths
A club records 86 people at one meeting and 90 at the next. Imagine a bar chart that starts both bars at 80. The visible lengths are 6 and 10 units. Those lengths represent attendance above 80, rather than the full attendance totals.
Read the labels, units and baseline before judging the comparison. The attendance increased by 4 people. Changing where the bars begin cannot change that count. A shared zero baseline and a linear scale let bar-length ratios represent value ratios.
Check tick spacing too. On a linear scale, equal distances represent equal numerical changes. Equally spaced labels of 0, 20, 40 and 200 would break that rule: the last change is much larger than the others.
A nonzero axis is not automatically deceptive. A line graph may use one to make small changes easier to inspect. Compare the actual values with the claim being made, rather than deciding from the graph's shape alone.
Check what the chart's shapes represent
A line often helps show change over time. Bars compare categories, while a pie divides one total into shares. Select the form that answers the question you are asking.
Imagine a budget with half the money assigned to food, a quarter to travel and a quarter to equipment. An accurate pie gives food half the circle and each other category a quarter. Labels or a key connect the portions to the categories. Adding more categories may make close shares harder to compare, but the number of slices alone does not make their sizes inaccurate.
Resized pictures need another check. If an icon doubles in height and width, its area becomes four times as large: two times two is four. If the value it represents only doubled, the area can exaggerate that increase. Read the values and find out what feature of the picture is meant to represent them.
Compare the crop with its source image
Imagine a cropped photograph showing a queue at a ticket window. The full photograph includes several nearby windows with no queue. The close view supports the observation that people waited at one window. It does not establish that every window had a queue.
When the larger image is available, identify what the crop removed. Explain which claim the missing details affect. The wider photograph still records only one view at one time; it cannot reveal everything elsewhere in the building or earlier that day.
A photograph also leaves questions about causes. A picture of a damaged fence shows damage. It does not by itself establish whether a storm, a collision or another event caused it. Look for evidence connecting the damage to a particular explanation.
Missing context is a reason to limit a conclusion. It does not prove that the opposite conclusion is true.
Explore more
2 · Watch
Take it off screen
Where it sits
Optional connections
This opens up
Where this leads
8 questions wait behind this lesson, each with its answer explained. Every answer feeds the sky: stars light as they are learned, and dim when it is time to come back.