Inference Question challenges

A valid inference must be true, based on the stimulus. Sometimes, it is directly stated in the text or immediately present in the visual data.

Other times, you can safely infer a statement because:

• It restates or summarises the text or data.
• It is unstated but very close to the surface meaning of the text.
• It can be confirmed with basic maths using the numbers and terms in the data.

If you can safely infer a statement, the answer is Yes.

An invalid claim is anything that is not supported by the stimulus. Sometimes, a statement is invalid because it directly contradicts the text or data.

Other times, an invalid statement could be true or false because:

• It is tangential to words or ideas in the text, but can’t be proven or disproved.
• It is a broader or narrower idea that cannot be clearly compared to the stimulus.
• It relies on numbers or ideas that go beyond the data provided in the stimulus.
• It combines words or ideas from the stimulus in a way that does not necessarily follow in all cases, possibly because the ideas are not linked in the stimulus.

Invest your efforts in improving your speed in evaluating the individual statements in a DM Inference question.

You have about 1 minute to answer each DM question under normal pacing.

If you use the Medify approach for five-part question triage, you can potentially bank several extra minutes for Syllogism and Inference questions after completing all the multiple-choice questions as fast as possible.

Even so, Test Day pressures can result in uneven performance. Most students report that Inference statements are broadly simpler and faster than Syllogism statements, especially with a text-only stimulus.

Refine your Inference skills, and you will have an opportunity to make up time in the DM section. If you have good pacing overall, you might have extra time for Syllogisms or other flagged questions.

Watch out for statements that look similar to the initial text, but with a slight shift in the terms or the overall meaning.

These statements are often wrong answer traps designed to make you select Yes without realising that the statement misaligns with the stimulus.

The logical approach is similar to a Verbal statement that seems True at first glance, but on further reflection is actually Can’t Tell.

Be vigilant and check logic as well as wording to see if a statement matches the meaning of the stimulus. If it does, it’s Yes. If not, it’s No.

If a statement in an Inference question is excessively wordy or complicated, don’t hesitate. Skip it and continue to evaluate other statements.

You can return to tough statements after you select Yes or No for shorter, simpler statements in the same question. Sometimes, those faster statements will help you see the reasons why the tougher ones are Yes or No.

Resist the urge to look at a table, graph or other visual data as soon as it appears.

Complex visual data is a time-wasting trap. You get marks from the statements, not the data.

Be rigorous in your Inference question technique:

• Read a statement so you know what you need to find in the data.
• Check the visual data to see if it has the details or data points you need.
• If not, quickly drag and drop No and move on.
• If you can eyeball or estimate the data points and see a match, select Yes and move on.
• If you need to calculate to confirm, use the calculator as fast as possible.
• Repeat the steps with the next statement.

If a table or graph is elaborate or challenging:

• Use your fingers as a guide to check you are reading the correct row, column, bar or line, or to check against axis labels and guidelines.
• Put a finger on each cell or data point to avoid your eyes drifting or misreading a value from an adjacent cell or data point.
• Try to compare numbers in your head – it may be sufficient to see it’s Yes or No.
• Calculate when needed – but think first to ensure you use the right steps and numbers.

It’s highly unlikely that an official DM Inference question will have 5 Yes or 5 No statements. There are a few official examples with 4 Yes or 4 No, so this is possible.

However, it’s most common by far to have 3 Yes and 2 No or 2 Yes and 3 No. Don’t fret if you select just 1 Yes or No, but you will likely have 2 or 3 of each in most Inference questions.

If you want to answer all No in a single question, it is likely you are missing something.

If you find that no statement is a Yes:

• Take a moment to check the text or data.
• Flag for review and come back to the question on a final pass through the DM section.

The same approach applies if you want to answer all Yes, though there has been one official Inference question with 5 Yes statements. It’s possible, if quite unlikely.

Each DM Inference question is worth two marks.

• You get two marks if you get 5 out of 5 statements correct.
• You pick up one mark if you answer 4 out of 5 statements correctly.
• You get no marks if you get 3 or fewer statements correct.

You must achieve good pacing through the DM section to give yourself the chance to earn both marks on most, if not all, of the Inference questions.

You can earn two marks on every DM Inference question regardless of difficulty.

A question with a text-only stimulus and fairly brief statements is worth two marks.

A question with complex visual data and relatively long statements is worth two marks.

If you find you struggle to finish text-only Inference questions, you may benefit from applying the VR scanning for key words technique. Remember that Inference questions usually have just a single text paragraph, or they will tend to have visual data with shorter text above or below it.

This means there is less text to check for support for each statement and select Yes when you see it’s a match. And if a statement contradicts the stimulus text, or is impossible to confirm as a match because it relies on outside information, select No and move on to the next statement.

You may improve your DM pacing if you triage across Inference questions:

• First, answer questions with only text – they tend to be much quicker.
• Then, attempt questions with graphs, charts or tables – they are more likely to require calculations and could take more time.

If you run out of time, it is better to rush or guess blindly on a question that would be harder to finish anyway. Do so with the confidence that you locked in marks in the other questions.

You can increase speed and accuracy in Inference questions if you triage the statements in each question:

• Always read the first statement before you look at the stimulus and try to answer it fairly quickly.
• If the first statement is very long or hard to understand, it’s better to skip it and start instead with the second statement.
• After you answer one statement, pick the simplest statement from the remaining four statements and attack it next.
• Continue in this manner until you answer all five statements, or until you decide to flag for review and come back to the final, hardest statements.

Train yourself to leave the hardest statement for last in every Inference question.

Think about what to do if you attempt an Inference question, then decide to leave one or two statements unanswered and flag for review.

Remember, the Review screen at the end of the DM section will:

• Indicate whether a question is flagged, incomplete or both.
• Allow you to select an individual question to review.
• Let you review only flagged questions or review only incomplete questions.

If you are sure you will have time to return to the question, it may be more efficient to leave statements blank. This way, you can quickly see the one(s) you must finish.

If you are not sure of any extra time, you might make a best guess for those statements, then make a note in your notebook, such as 18-3 (for question 18, 3rd statement).

If an Inference question includes visual data on your second pass through the section:

• Take a quick glance at the statements to get a broad sense of their complexity. This means, very roughly, how much work you must do to answer Yes or No.
• If it looks like lots of calculations are required in multiple statements, it’s very likely the best option is to skip and flag for review. Do the question at the very end of your second pass through the section.
• If it looks relatively simple and like you might be able to answer without calculating, you can answer it now or do it after text-only questions, but before the question(s) with lots of calculations.

Again, this is very much a judgement call. Try to refine this skill in your DM practice.