10.5 Data Interpretation Sets

10.5 Data Interpretation Sets

Quick Answer: DI sets are 2-3 questions sharing one data source (table, bar/line/pie chart, or scatterplot). Read the data first - title, units, scale, totals, footnotes - then attack the questions. Unlike geometry figures, DI graphics ARE drawn to scale, so estimation from axes is allowed.

What a DI Set Looks Like

A DI set begins with a data presentation: a table, a bar chart, a line graph, a pie chart, or occasionally a scatterplot or a compound chart. Below the data sit 2-3 numbered questions that all reference that same data. Each question is a standard GRE question type:

• Quantitative Comparison - compare two quantities derived from the data.
• Multiple Choice (single answer) - 5 choices, exactly one correct.
• Multiple Choice (multiple answer) - 3 choices, 1-3 correct, no partial credit.
• Numeric Entry - compute and type the value.

ETS reports that DI accounts for a meaningful chunk of the Quant section - plan for 3-5 DI questions in each of the two scored sections, often grouped into 2 sets per section.

The DI Reading Protocol (Do This First)

Before reading a single question, spend 30-60 seconds with the data:

1. Title. What is being measured? "Monthly Sales by Region, 2024" tells you the unit (sales, probably dollars), the granularity (monthly), and the scope (by region, year 2024).
2. Units. Are values in dollars, thousands of dollars, millions? Percent or percent change? Index points? The unit label on the axis or table header is the single most-tested DI fact.
3. Scale. Does the bar chart start at 0 or at a non-zero baseline? Is the y-axis logarithmic? A truncated y-axis exaggerates differences.
4. Legend. Multi-series charts list which color or pattern maps to which category. Memorize it; you will reference it on every question.
5. Totals and rounding notes. Footnotes such as "Figures are rounded to the nearest thousand" or "Totals may not sum due to rounding" explain why your arithmetic might be off by a few units.
6. Time periods. Note the year range, especially if a question asks about change over time.

Why the Upfront Read Pays Off

DI sets are the one place on the GRE where reading time is leveraged. A 45-second scan saves 20-30 seconds on each of the next 3 questions, because you will already know the units, the scale, and where to find the relevant row or bar. Skipping the scan to "save time" is a false economy - you will re-read the chart three times.

Estimation vs. Exact Computation

DI graphics are drawn to scale. This is the opposite of geometry figures (Section 10.7), which are NOT drawn to scale. Consequences:

• For bar and line charts, you can estimate a value by reading the axis - useful for eliminating absurd answer choices.
• For pie charts, angle measures correspond to proportions, so a 25% slice is a 90-degree wedge.
• For tables, you must compute; there is no visual estimate.
• When a question says "approximately," lean on estimation: the answer choices will be far enough apart that visual reading suffices.
• When a question says "exactly" or asks for a Numeric Entry, compute with the calculator.

Percent vs. Absolute Trap

DI questions love to blur the line between percentages and absolute counts.

• "What percent of total sales came from Region A?" -> divide Region A sales by total sales.
• "What was the percent increase in Region A sales from 2023 to 2024?" -> divide (2024 - 2023) by 2023.
• "How many more units did Region A sell than Region B?" -> subtract absolute counts, not percents.
• "Region A's share rose 10 percentage points" -> if 2023 share was 30%, 2024 share is 40%, NOT 40% growth. Percent growth would be (40 - 30) / 30 = 33.3%.

Sample DI Set

The table below shows 2024 monthly sales (in thousands of dollars) for two product lines at a single retailer.

Footnote: Sales are rounded to the nearest thousand dollars.

Question 1 (Numeric Entry): What was the total sales for Alpha across the six months, in thousands of dollars? (Enter as an integer.)

Sum the Alpha column: 32 + 35 + 40 + 38 + 45 + 50 = 240. The unit is already thousands of dollars, so enter 240. A common error is to multiply by 1,000 to convert to dollars, but the table header already states the unit. Reading the header during the upfront scan prevents that mistake.

Question 2 (Multiple Choice, single answer): What was the percent increase in Beta sales from January to June? Round to the nearest whole percent.

• January Beta = 18; June Beta = 36.
• Change = 36 - 18 = 18.
• Percent increase = 18 / 18 = 1.00 = 100%.
• Answer: 100%.

A frequent trap is computing 36 / 18 = 2 and reporting 200% - that is the ratio of June to January, not the increase. Percent increase always divides the difference by the original (January) value, not the new value.

Question 3 (Quantitative Comparison): Compare Quantity A: the median monthly Alpha sales, with Quantity B: 39.

• Alpha sales sorted: 32, 35, 38, 40, 45, 50 (six values).
• Median of an even count = average of the two middle values: (38 + 40) / 2 = 39.
• Quantity A = 39; Quantity B = 39. The answer is C - the two quantities are equal.

If Alpha had five values, the median would be the middle one; with six, average the third and fourth. A common QC trap is to take the unsorted middle value (40) instead of sorting first.

Pacing and When to Skip

DI sets reward careful first reads but punish obsessive precision. Plan for 90 seconds per DI question on average - more if the chart is dense, less if the question is a simple read. If a set has a question that requires multi-step percent change across non-adjacent months and you are short on time, mark it for review and complete easier questions first. The first question of a set is often the easiest (a direct read); the last is often the hardest (a multi-step computation).

Strategic DI Tips

• Scan the data first; always. The 45-second scan is non-negotiable.
• Write down units on scratch paper before the first question. "Alpha and Beta in $ thousands" prevents a class of errors.
• Estimate when the question says "approximately." The choices will be far apart; visual reading from the bar height suffices.
• Compute when the question asks for an exact value or a Numeric Entry. Use the calculator; do not trust visual estimation for the final digit.
• Re-check percent-change direction. Increase divides by the original (earlier) value, not the new value.
• Sort before taking a median. The chart's left-to-right order is chronological, not numerical.
• Note any footnote. Rounding notes explain small mismatches; a missing footnote is a red flag that exact computation is expected.