Study guide
Beyond Data Sufficiency, the Data Insights section includes four more question types that test how well you synthesize information presented as text, tables, and graphics under a shared 45-minute, 20-question section limit, and a basic on-screen calculator is permitted throughout Data Insights. This chapter covers Multi-Source Reasoning, Table Analysis, Graphics Interpretation, and Two-Part Analysis, the remaining four of the section's five official question types.
Multi-Source Reasoning
Multi-Source Reasoning questions present information across two or three tabs, which may contain a text passage, a table, a chart, or an email-style exchange, and a set of two to three questions follows that require you to synthesize details from more than one tab at once. For example, one tab might describe a company's shipping policy in prose, a second tab might contain a table of order weights and destinations, and a question could ask which of several orders would qualify for expedited shipping, a determination that requires reading the policy's numeric thresholds from the text tab and then checking each order's weight against those thresholds in the table tab. A common trap is answering from only one tab because it contains the most obviously numeric information, while missing a qualifying condition stated only in the prose tab, such as an exception for orders shipped to certain regions. Some Multi-Source Reasoning questions are structured as a series of true/false or yes/no statements about a single scenario, where each statement must be evaluated independently against all the available tabs, and partial credit is not given; every part of a multi-statement question must be answered correctly to receive credit for that question. The efficient approach is to skim all tabs once for their general content and organization before reading the first question closely, so you know where to look back for specific figures or conditions without rereading everything from scratch each time.
Table Analysis
Table Analysis presents a single data table, similar to a spreadsheet, that you can sort by any column using an on-screen control, and questions typically ask you to evaluate a series of statements as true or false, or as would/would not satisfy a stated condition, based on the sorted data. For example, given a table listing employees with columns for department, years employed, and annual salary, a question might ask whether it is true that the employee with the longest tenure in the marketing department also has the highest salary in that department; sorting the table first by department and then by tenure within that filtered view makes this comparison direct rather than requiring you to scan the unsorted table row by row. As with Multi-Source Reasoning, Table Analysis questions are often presented as multiple linked statements about the same table, each requiring its own true/false judgment, with credit given only when every statement in the set is answered correctly. A frequent error is applying a filter mentally rather than actually re-sorting the table, since scanning an unsorted table by eye under time pressure is much more error-prone than using the built-in sort function to bring relevant rows together. Another frequent trap involves statements using words like only, all, or at least, which require checking every relevant row in the table rather than a few representative ones, since a single counterexample row is enough to make an all or only statement false.
Graphics Interpretation
Graphics Interpretation questions present a single graphic, such as a scatter plot, a bar chart, a pie chart, a line graph, or a statistical distribution curve, and ask you to complete one or more sentences by choosing from a small dropdown menu of options for each blank, based on reading values or relationships directly from the graphic. Consider a scatter plot of advertising spend on the x-axis against monthly sales on the y-axis for twenty stores, with a labeled trend line sloping upward; a Graphics Interpretation question might ask you to complete the sentence the store with the (highest / lowest) advertising spend had monthly sales of approximately (300 / 500 / 700) thousand dollars, requiring you to locate the correct point and read its approximate y-value. Because these questions are graded on precise reading of the visual, look for axis labels, units, and scale carefully before answering; a bar chart with a y-axis that does not start at zero can make differences between bars look larger than they actually are in absolute terms, and misreading the scale is a common source of error. For a statistical curve, such as a bell-shaped distribution, questions may ask about the mean, the median, or the general shape of the distribution, for instance whether it is symmetric or skewed toward one tail; recognizing that a longer tail on one side pulls the mean toward that tail, while the median is less affected, connects this topic directly to the statistics content covered in Quantitative Reasoning.
Two-Part Analysis
Two-Part Analysis presents a single scenario, often involving a system of relationships, a word problem, or a short argument, followed by a table with two answer columns and several answer-choice rows; you must select exactly one row for each column, and the two parts may be related quantitatively, verbally, or a mix of both. A typical quantitative Two-Part Analysis item might describe two variables constrained by two equations and ask you to select, from a shared list of numbers, which value satisfies one equation for the first part and which value satisfies the other equation for the second part; because both parts often draw from the same answer list, solving one part can sometimes eliminate options for the other. A verbal Two-Part Analysis item might present a business scenario and ask you to identify, from a list of statements, which one is most likely a cause of a described problem for the first part and which one is most likely a solution consistent with given constraints for the second part. Unlike Multi-Source Reasoning and Table Analysis, where linked statements are graded as a single all-or-nothing set, Two-Part Analysis still requires both parts to be answered together as one question, so treat the pair as inseparable: read both column headers and the full scenario before committing to either selection, since a choice that looks correct for one column in isolation may not fit once the second column's requirement is factored in.
Key terms
- Multi-Source Reasoning
- — A Data Insights question type presenting information across two or three tabs (text, table, and/or graphic) that must be synthesized to answer linked questions.
- Tab
- — One of the two or three separate panels of information, such as a passage, table, or chart, provided in a Multi-Source Reasoning prompt.
- Table Analysis
- — A Data Insights question type presenting a single sortable, spreadsheet-style table, with questions evaluated by sorting and filtering the data.
- Sortable column
- — A column in a Table Analysis table that can be reordered using an on-screen control to bring relevant rows together for comparison.
- Graphics Interpretation
- — A Data Insights question type asking you to complete sentences using dropdown selections based on values or relationships read from a chart or graph.
- Trend line
- — A line drawn through or near a set of scatter-plot points to indicate the general direction of the relationship between two variables.
- Skewed distribution
- — A statistical distribution with a longer tail on one side, which pulls the mean toward that tail more than it affects the median.
- Two-Part Analysis
- — A Data Insights question type requiring two related selections, one from each of two answer columns, based on a single shared scenario.
- Linked statement set
- — A group of true/false or yes/no statements attached to one Multi-Source Reasoning or Table Analysis prompt, graded together as a single all-or-nothing unit.
- On-screen calculator
- — A basic calculator tool made available during the Data Insights section, unlike Quantitative Reasoning, where no calculator is permitted.
Exam tips
- Skim every tab in a Multi-Source Reasoning prompt once before answering, so you know which tab holds which kind of detail and can look back efficiently.
- In Table Analysis, use the actual sort function on the relevant column rather than scanning the unsorted table by eye, especially for statements using only, all, or at least.
- In Graphics Interpretation, check the axis labels, units, and starting value of the scale before reading any specific point, since a non-zero baseline can distort visual comparisons.
- In Two-Part Analysis, read both column headers and the full scenario before selecting either answer, since the two parts are often interdependent.
- Remember that linked statement sets in Multi-Source Reasoning and Table Analysis typically require every statement correct for credit, so do not rush past any single statement in the set.