Session 08  ·  Student Handout  ·  Phase 2 · Information & Ideas

Command of Evidence — Graphs & Data

Graph questions are not math questions — they're reading comprehension questions with a chart. The number skills required are minimal. The reading skills are everything.

Reading, Not Math Information & Ideas Session 08 · Phase 2
01 Graph Anatomy — What to Read in 20 Seconds
What Makes Graph Questions Different
Unlike textual evidence questions, graph questions pair a short passage with a visual data source — a bar chart, line graph, or table. Your job is identical to Session 7: find evidence that directly proves a specific claim. The only difference is that the evidence is in the chart, not in a quotation. The same rules apply: topic-relevant data is not the same as claim-relevant data.

Before reading any question, spend 20 seconds reading the graph. Hit these four elements in order:

Anatomy of a Graph Question — Example Bar Chart
Average Weekly Park Visits by City Type (2022)
543210
3.2 Dense Urban
4.1 Suburban
2.2 Rural
4.5 Near-Urban
City Type
① Title
"Average Weekly Park Visits by City Type" — tells you what is being measured and how it is grouped. Always read this first.
② Axes & Units
Y-axis = average weekly visits (not percentage, not dollars). X-axis = city type categories. Units tell you what the numbers mean.
③ Trend / Pattern
Suburban and near-urban residents visit most; rural visit least. The pattern is what answer choices will test — not the exact numbers.
④ Source Note
Who collected this data and when? This matters for evaluating claim scope — a 2022 US survey does not prove global or historical trends.
02 The 4-Step Graph Question Method
Step 1
Read the Graph
Title → axes → units → trend. 20 seconds. Do this before reading the question.
Step 2
Isolate the Claim
What specific claim does the question or passage make? Write it in one sentence before reading the choices.
Step 3
Find the Proof
Which specific data point or trend in the graph directly proves that claim? Numbers must match the claim exactly.
Step 4
Match to Choice
Find the answer choice that accurately states what you found in step 3. Eliminate any choice that adds, omits, or distorts the data.
The Most Important Rule for Graph Questions
03 Correlation ≠ Causation — The Most Common Graph Trap

A graph can show that two things move together (correlation) without proving that one causes the other (causation). This distinction generates more wrong answers on graph questions than any other single trap.

✗ Causation Claim (Wrong)
"The data shows that increased park access causes lower stress levels in urban residents."
The graph shows a correlation between park proximity and reported stress. It cannot prove causation — something else (income, neighbourhood quality) might explain both.
✓ Correlation Claim (Correct)
"The data shows that residents with park access within half a mile reported lower average stress scores than those without such access."
Accurately states what the graph shows — a pattern — without claiming to explain why the pattern exists.
Automatic Elimination Rule
Any answer choice that uses the words causes, leads to, results in, proves that, demonstrates that X is responsible for Y — when the evidence is only a graph or dataset — is almost certainly wrong. Graphs show patterns and correlations. They do not establish causation unless the passage or question explicitly states that a controlled experiment was conducted.
04 Worked Examples
Worked Example 1 Bar chart · Claim-matching · Natural Science
Natural Science
A researcher studying urban ecology argues that city residents who live within half a mile of a public green space report meaningfully lower stress levels than those who live farther away.
AVERAGE REPORTED STRESS SCORE BY DISTANCE FROM NEAREST GREEN SPACE
(Scale: 1 = Very Low Stress, 10 = Very High Stress · n = 1,240 urban residents)
Avg. Stress Score (1–10)
86420
3.2 < 0.5 mi
5.1 0.5–1 mi
6.5 1–2 mi
7.6 > 2 mi
Distance from Nearest Green Space
Source: Urban Wellbeing Survey, 2023
Step 1 — Read the graph: Title = stress scores by distance from green space. Y-axis = stress score (1–10). Trend = stress increases as distance increases. Closest group (under 0.5 mi): score 3.2. Farthest group (over 2 mi): score 7.6.
Step 2 — Isolate the claim: "Residents within 0.5 miles report meaningfully lower stress than those farther away."
Step 3 — Find the proof: The 0.5 mi group scores 3.2 vs. 7.6 for the 2+ mi group — a difference of 4.4 points on a 10-point scale. That is "meaningfully lower."

Which choice most effectively uses data from the graph to support the researcher's argument?

ACity residents who visit green spaces regularly cause their stress levels to decrease significantly over time.— Causation claim. Graph only shows correlation. Eliminate.
BMost urban residents in the study lived more than one mile from a green space.— Topic-relevant but proves nothing about the stress claim.
CResidents living within half a mile of a green space reported an average stress score of 3.2, compared to 7.6 for those living more than two miles away — a difference of more than four points on a ten-point scale.✓ Direct, accurate, claim-proving data from the graph.
DAccess to green spaces is the most important factor determining stress levels among urban residents.— Outside the graph. "Most important factor" is a causal and comparative claim the data doesn't support.

Key: C is the only choice that accurately states specific data values from the graph and connects them directly to the claim. A is a causation trap. B is topic-relevant but proves nothing about stress levels. D adds a comparative claim ("most important factor") that the graph cannot support.

Worked Example 2 Line graph · Correlation trap · Economics
Social Science / Economics
An economist argues that the minimum wage increases enacted in several U.S. cities between 2014 and 2020 did not reduce employment in the restaurant industry in those cities.
RESTAURANT INDUSTRY EMPLOYMENT INDEX — CITIES WITH MINIMUM WAGE INCREASES vs. COMPARISON CITIES
(Index: 2014 = 100)
Employment Index (2014 = 100)
11511010510095
2014 2015 2016 2017 2018 2019 2020
Cities with wage increases
Comparison cities
Source: Bureau of Labor Statistics Regional Employment Data, 2021
Step 1 — Read the graph: Both city types show increasing employment (both lines go up). Cities with wage increases grew from index 100 to ~113. Comparison cities grew from 100 to ~108. Neither line goes down.
Step 2 — Isolate the claim: Minimum wage increases did NOT reduce restaurant employment.
Step 3 — Find the proof: Cities with increases saw employment RISE (not fall), and actually rose more than comparison cities.

Which choice most effectively uses data from the graph to support the economist's argument?

ARestaurant employment in cities with minimum wage increases fell sharply after 2016 before recovering by 2020.— Directly contradicted by the graph. Employment rose consistently.
BMinimum wage increases cause restaurant owners to cut staff because labour costs rise faster than revenue.— Causation claim, and directly contradicts the graph trend.
CRestaurant employment in cities with minimum wage increases rose from an index of 100 in 2014 to approximately 113 in 2020, while comparison cities rose only to approximately 108 — indicating no employment decline in wage-increase cities.✓ Accurate data, correct direction, directly proves "no reduction."
DBoth city types showed similar levels of restaurant employment throughout the study period.— Misreads the graph. Wage-increase cities actually outperformed comparison cities by 2020.

Key: Always read the direction and magnitude of changes, not just whether there is a difference. A wrong answer can get the direction wrong (A), add a causation claim (B), or misread the comparison (D). C is the only choice that accurately reads both lines and connects the data to the claim.

Session 8 — The Three Rules
Ready to practice?

14 questions — each one includes an inline bar chart, line graph, or data table. Guided practice walks you through the 4-step method. Timed section includes questions with two data sources plus a passage.

Open Session 8 Exercises →