Visual traps: scale illusions. Math traps: % vs. absolute, wrong base. Logic traps: correlation ≠ causation, "all" vs. "some." Timing traps: level vs. rate, cumulative vs. period.
GMAT DI questions are carefully designed to exploit specific cognitive shortcuts. Knowing the trap types in advance lets you recognize them in real-time and avoid common errors.
A country with 5% market share of a $10B market ($500M) is larger than one with 20% of a $2B market ($400M). Always compute absolute values when totals differ.
A higher average doesn't mean a higher total if the count differs. 10 workers averaging $50K = $500K total. 20 workers averaging $30K = $600K total.
% change uses the ORIGINAL value as denominator. Price goes from $80 to $100: change = 20/80 = 25%, not 20/100 = 20%.
A scatter plot showing two variables increasing together does NOT prove one causes the other. Always flag when answer choices claim causation from correlation data.
A statement saying "all companies with X also have Y" is False if even ONE company has X but not Y. "Some" only needs one example. Get the quantifier right.
If the data covers 10 companies in North America, you cannot conclude anything about global trends. Scope of data limits scope of conclusions.
| Trap | The Mistake | The Fix |
|---|---|---|
| Different time periods | Comparing Q1 to Q4 as if same base | Ensure periods are the same length and base |
| Level vs. rate of change | Steepest slope = highest value | Slope = rate of change, not absolute level |
| Base year selection | Choosing a conveniently low/high base | Check what reference year was used for indexing |
| Cumulative vs. period | Treating a cumulative total as a period value | Read axis label: "YTD" or "annual" or "per month" |
The biggest value in one metric (e.g., employees) doesn't make it the "best" company on another metric.
Truncated y-axis makes small differences look huge. Always check the baseline before visual comparisons.
A 50% growth rate and a 50% market share are totally different statistics.
Line graphs imply smooth change but actual values between labeled points may not be linear.
In a chart with 3+ lines, verifying which line is which (via legend) is essential before every reading.
If a table shows data for 5 of 500 companies, conclusions about the whole industry are not supported.
At least 50% includes exactly 50%. More than 50% excludes it. One extra row can flip a True/False.
In scatter plots with two unlabeled axes, the x-axis is conventionally the independent (cause) variable — but confirm with the context.
If the true answer is 50.1% and you round to 50%, a "more than half" True answer becomes False. Use precise arithmetic for borderline cases.
This answer is correct ONLY when the given data is genuinely insufficient. Never choose it just because you find the question hard.
A chart shows that Company A grew revenue by $50M while Company B grew by $20M. A student concludes "A grew faster than B." This conclusion is:
A scatter plot shows a perfect negative correlation between hours of TV watched per day and GPA. A teacher claims: "Watching less TV will improve your GPA." This claim is:
A bar chart shows 2023 vs. 2022 values. Company X: 2022=$200M, 2023=$220M. The percentage increase is:
A Table Analysis statement says: "All products with a price above $100 have a customer rating above 4.0." The most efficient way to test this is:
A pie chart for Year 1 shows Company A = 40% of a $500M market. A pie chart for Year 2 shows Company A = 35% of a $700M market. A says "A's revenue fell because its share decreased." This is:
A line graph shows "Cumulative Revenue" (not annual). The value goes from $100M in Year 1 to $250M in Year 3. Annual revenue in Year 3 alone cannot be determined from this graph unless:
A table has 10 companies. A statement says "More than half of the companies with revenue above $1B also have profit margins above 20%." The statement is TRUE if:
A GI question shows a bar chart with y-axis from 50 to 100 (truncated). Bar A is at 75, Bar B is at 90. Visually, B appears twice as tall as A because of the truncation. The actual ratio of B to A values is approximately:
An MSR question provides two tabs: Tab 1 shows a sales report, Tab 2 shows a customer survey. Neither tab mentions production costs. A question asks "What was the production cost per unit?" The correct answer is:
A TPA question asks you to select "an event that preceded the outcome" (Column 1) and "a mechanism that explains the outcome" (Column 2). Three students pick the same Column 1 answer but different Column 2 answers. This shows:
Truncated axes, 3D distortion, and scale mismatches are the most common visual traps.
% says nothing about actual size. Always multiply by the total when comparing across categories.
No scatter plot, table, or chart can prove causation. Never choose answers that claim it.
A steep slope shows fast change, not necessarily high values. Cumulative and period values are different.