Data Analysis is the backbone of GRE Data Interpretation. In every DI set you will encounter a shared data display — a table, graph, or chart — followed by questions that test your ability to read values, compute statistics, compare categories, and interpret trends. Unlike standalone quantitative questions, DI data analysis demands that you extract the right numbers from a visual before doing any math. Below you will learn the eight question stem patterns that appear, work through two interactive examples step by step, and practice with six guided questions pulled from realistic GRE-style data sets.
What Are DI Data Analysis Questions?
Data Interpretation questions on the GRE arrive in sets: a shared data display (table, line graph, bar chart, circle graph, or frequency distribution) followed by two to five questions. Data Analysis questions are those that require you to read values from the display, compute statistics such as the mean or median, calculate rates of change, compare categories across time periods, or draw conclusions about trends. They test whether you can work fluently with data presented in visual and tabular formats.
Unlike DI questions that require algebraic reasoning or geometric interpretation, data analysis questions are grounded in arithmetic and statistical thinking. The computations themselves are usually straightforward — the difficulty comes from reading the correct series on a graph, counting intervals versus data points accurately, and selecting the right statistic (mean vs. median, total change vs. average change). Getting these details wrong is exactly how the GRE generates its distractor answers.
Frequency note: Data Analysis is the most common domain tested in DI sets. Nearly every DI set on the GRE includes at least one data analysis question, and many sets are entirely focused on this domain. Expect to see data analysis skills tested in the majority of your DI encounters on test day.
8 Patterns You'll See
Nearly every DI data analysis question matches one of these eight stem patterns. Recognizing the pattern tells you what to compute and which data to extract from the display.
1
Ratio of A to B
Compute a ratio between two values read directly from the data display. Requires identifying the correct rows, columns, or data series for both A and B.
2
Percent of Group
Determine what percent one value is of a total. Common in table-based sets where you compute a category's share of a column total or a row total.
3
Which Must Be True (Data Statements)
Evaluate multiple statements about the data and determine which ones are necessarily true. Requires careful verification of each statement against the display.
4
For Which Categories Did a Condition Hold
Scan across years, categories, or data points to identify all entries where a stated condition is satisfied. Requires systematic checking, not estimation.
5
Average Rate of Change
Compute intervalsfinal−initial. The classic trap: using the number of data points instead of intervals as the divisor.
6
Percent Greater or Less Than
Compare two values from the display and express the difference as a percentage of one of them. Pay attention to which value is the base (denominator).
7
Fraction of a Subset Satisfying a Condition
Combine data from multiple rows or columns, then compute a fraction. Often involves weighted combinations where different subgroups have different sizes.
8
Hypothetical Data Correction
If one data point is changed, how does a statistic (usually the mean) change? The shortcut: change in mean = nΔvalue.
How to Solve DI Data Analysis Step by Step
Apply these six strategies in order to handle any DI data analysis question accurately and efficiently.
In a multi-line or multi-bar graph, make sure you are reading the correct series before extracting any numbers. Check the legend carefully. A common source of wrong answers is reading the exports line when the question asks about imports, or pulling data from the wrong column in a multi-column table.
The phrase "the years 2013 through 2017" includes both endpoints, giving you 5 years (2013, 2014, 2015, 2016, 2017). Write them out if you are unsure. On the GRE, "between 2013 and 2017" typically also includes both endpoints. Miscounting by one changes your average and often matches a distractor.
The rate of change from 2012 to 2017 spans 5 intervals, even though there are 6 data points (2012, 2013, 2014, 2015, 2016, 2017). Dividing by 6 instead of 5 is one of the most common GRE errors. The wrong answer generated by this mistake is almost always offered as a choice.
With n data points, the median is at position (n+1)/2. Locate this position in the cumulative frequency. For example, with 75 data points the median is the 38th value. Accumulate frequencies from the lowest category until you reach or pass position 38 to determine which category contains the median.
Approximate values by noting where a bar or line falls relative to labeled gridlines. If a bar falls halfway between 4 and 6 on the axis, the value is approximately 5. For line graphs, find the year on the horizontal axis, trace vertically to the line, then trace horizontally to the vertical axis to read the value.
After computing an answer, scan the answer choices before selecting one. If your value does not match any choice exactly, recheck whether you used the right data series, the right base for a percent calculation, or the right divisor for a rate. Distractor choices are designed to match the results of the most common errors, so a close match is a warning sign, not a confirmation.
Pro tip: Use the answer choices to calibrate your precision. If the choices are $6.4 billion, $6.6 billion, $6.8 billion, $7.0 billion, and $7.2 billion, your calculation must be precise. If the choices are $200 million, $400 million, and $600 million, a rough estimate will suffice. Let the spread of the answer choices guide how carefully you read the graph.
Worked Example: Average Rate of Change
Work through each step below. You must answer each mini-challenge correctly to unlock the next step. If you get stuck, a second wrong attempt will reveal the answer so you can keep going.
Interactive Walkthrough0/7 steps
Computing Rate of Change from a Production Table
The table below shows annual maize production (in millions of tonnes) for the years 2012 through 2017.
Annual Maize Production (millions of tonnes)
Year
2012
2013
2014
2015
2016
2017
Production
273
351
361
345
385
371
Which of the following is closest to the average rate of change, in millions of tonnes per year, for maize production from 2012 to 2017?
1
Step 1: Identify the endpoints
For the average rate of change from 2012 to 2017, which two values do you need?
2
Step 2: Read the initial value
3
Step 3: Read the final value
4
Step 4: Compute the total change
5
Step 5: Count the intervals
6
Step 6: Compute the average rate of change
7
Step 7: Identify the trap answer
Worked Example: Median from a Frequency Distribution
This example teaches you how to find the median when data is presented as a frequency distribution table rather than as individual values. Work through each step to build the cumulative frequency approach.
Interactive Walkthrough0/6 steps
Finding the Median from a Frequency Table
A survey of 75 consumers recorded how many gallons of milk each purchased last week. The frequency distribution is shown below.
Gallons of Milk Purchased Last Week (75 consumers)
Gallons
0
1
2
3
4
5
Consumers
14
23
18
11
6
3
Quantity A: The median number of gallons purchased. Quantity B: 1.5. Which is greater?
1
Step 1: Verify the total count
What is 14 + 23 + 18 + 11 + 6 + 3?
2
Step 2: Find the median position
3
Step 3: Cumulative frequency through 0 gallons
4
Step 4: Cumulative frequency through 1 gallon
5
Step 5: Locate the 38th value
6
Step 6: Compare to Quantity B
Practice Questions
Now apply what you learned. Each question has a step-by-step solution walkthrough — after you submit your answer, click through the solution one step at a time to compare against your own work.
StreamNet Inc. — Quarterly Internet Traffic and Revenue
Table 1: Quarterly Internet Traffic Volume (in petabytes) by Content Category for StreamNet Inc., FY2024
Content Category
Q1
Q2
Q3
Q4
Video Streaming
420
385
510
575
Music & Audio
145
160
155
180
Cloud Gaming
210
245
290
320
Social Media
330
310
345
360
News & Publishing
95
85
110
130
Table 2: Revenue per Petabyte (in thousands of dollars) by Content Category for StreamNet Inc., FY2024
Content Category
Q1
Q2
Q3
Q4
Video Streaming
12.5
11.8
13.2
14.0
Music & Audio
8.4
8.8
9.0
9.5
Cloud Gaming
18.0
17.5
19.2
20.5
Social Media
6.2
5.8
6.5
7.0
News & Publishing
15.0
14.2
15.8
16.5
Question 1 — Percent of Total (Pattern 2)
In Q4, the revenue from Cloud Gaming was approximately what percent of the total revenue from all five content categories combined?
For which content category did the revenue per petabyte increase by the greatest percent from Q1 to Q4?
Question 3 — Percent Increase of Revenue (Pattern 6)
From Q1 to Q4, the total revenue from Social Media (traffic volume multiplied by revenue per petabyte) increased by approximately what percent?
GlobalTech Corp — Revenue and Training Investment
Table 1: Average Revenue per Employee (in thousands of dollars) by Department at GlobalTech Corp, 2019-2023
Department
2019
2020
2021
2022
2023
Engineering
285
270
310
345
380
Marketing
195
180
205
230
255
Sales
245
215
260
290
320
Operations
150
140
165
185
210
Customer Support
120
110
135
155
175
Table 2: Annual Training Investment per Employee (in dollars) by Department at GlobalTech Corp, 2019-2023
Department
2019
2020
2021
2022
2023
Engineering
4,200
3,800
5,100
5,600
6,200
Marketing
2,800
2,500
3,200
3,500
3,900
Sales
3,500
3,100
4,000
4,400
4,800
Operations
2,200
1,900
2,600
2,900
3,300
Customer Support
1,800
1,600
2,100
2,400
2,700
Question 4 — Greatest Percent Increase by Department (Pattern 6)
From 2019 to 2023, which department experienced the greatest percent increase in average revenue per employee?
Question 5 — Rate of Change and Median (Pattern 5)
The median of the five departments' average annual rates of change in training investment per employee from 2019 to 2023 is closest to which of the following, in dollars per year?
Question 6 — Percent Exceeds (Pattern 6)
In 2023, the sum of the average revenues per employee across all five departments exceeded the corresponding sum for 2020 by approximately what percent?
Common Traps
Trap 1 — Mean vs. Median Confusion. The mean and median are different statistics and can give very different values, especially in skewed distributions. Read the question carefully to determine which one is being asked for. On the GRE, the wrong statistic is almost always offered as a distractor. If the question says "arithmetic mean," compute the average. If it says "median," sort and find the middle value.
Trap 2 — Rate of Change: Total vs. Average. The total change from 2012 to 2017 is simply Value_2017 minus Value_2012. The average annual rate of change divides that total by the number of intervals (5, not 6). These two quantities differ by a factor of 5, and each is offered as a wrong answer for the other. Always check whether the question asks for the total change or the average rate.
Trap 3 — Misreading "Percent Of" in Context. When a question asks "X was what percent of Y," be precise about what Y represents. If the question says "the production used for ethanol was what percent of the production not used for exports," the denominator is (100% - export%) of total production, not total production itself. Using the wrong denominator shifts your answer and typically matches one of the provided choices.
Recognition / When to Apply
Use this table to quickly identify which data analysis technique a DI question is testing based on its wording. Matching the question stem to the right approach saves time and prevents errors.
Question Stem Cue
Technique
Watch Out For
"What is the ratio of A to B?"
Read two values, divide
Using the wrong data series or wrong year
"What percent of [group] are [category]?"
Part/whole×100
Using the wrong total as the denominator
"What is the average rate of change?"
(Final−Initial)/intervals
Dividing by data points instead of intervals
"What is the median?"
Sort values, find middle position
Computing the mean instead of the median
"For which years did [condition] hold?"
Check each year systematically
Estimating instead of computing each value
"X is what percent greater than Y?"
(X−Y)/Y×100
Using X as the base instead of Y
"If [value] were corrected, how would the average change?"
Δmean=Δvalue/n
Reporting the raw change instead of dividing by n
"Which must be true?"
Verify each statement against the data
Assuming a statement is true without checking edge cases
The most reliable recognition signal for data analysis questions is that they ask you to compute or compare values directly from the data display without introducing unknowns or hypothetical variables. If the question asks you to simply read, compute, and compare, it is data analysis. If it introduces an unknown (such as a corrected value or a projected value), it may cross into algebra territory — but the underlying data reading skill is still data analysis.
Study Checklist
DI Data Analysis Mastery Checklist0/8 complete
Frequently Asked Questions
How frequently do Data Analysis questions appear in GRE Data Interpretation sets?
Data Analysis is the most common domain tested in DI sets. Nearly every DI set on the GRE includes at least one question that requires data analysis skills — reading values from graphs, computing averages, comparing categories, or interpreting trends. Many DI sets consist entirely of data analysis questions. You should expect to encounter this question type on every GRE administration.
What is the difference between mean and median on GRE DI questions?
The mean (arithmetic average) is the sum of all values divided by the count of values. The median is the middle value when the data is sorted in ascending order. For an even number of values, the median is the average of the two middle values. These statistics can yield very different results, especially in skewed distributions. The GRE frequently tests whether you know which statistic is being requested, and the wrong one is almost always among the answer choices.
How do I compute the average rate of change from a GRE graph?
The average rate of change equals (final value minus initial value) divided by the number of intervals. The critical detail is that the number of intervals is one less than the number of data points. For example, data from 2012 to 2017 includes 6 data points but spans only 5 intervals. Dividing by 6 instead of 5 produces a wrong answer that typically matches one of the distractor choices. Only the endpoint values matter for this calculation — you do not need the intermediate values.
Do I need to read every value from a graph to answer GRE DI questions?
Not always. Many questions require only specific data points. Before you begin reading the graph, identify exactly what the question asks for. Rate-of-change questions typically need only the endpoints. Median questions from frequency distributions need cumulative frequencies, but often you can stop once you have located the position. Reading only the necessary values saves time and reduces the chance of transcription errors.
What is the most common mistake on GRE Data Interpretation data analysis questions?
The single most common mistake is confusing the number of data points with the number of intervals when computing rates of change. If data spans from 2012 to 2017, there are 6 years (data points) but only 5 year-to-year intervals. Dividing by 6 instead of 5 produces an answer that is systematically too low, and this wrong answer is virtually always offered as a choice. A close second is computing the mean when the question asks for the median, or vice versa.