Part 2: Reading to Apply a Diagram
A chart of four near-identical options, an email that applies it to one person, and eight questions in eight minutes. The shortest text in the section, and the one most often misread as easy.
The part that looks like a break
After a 350 word letter, Part 2 appears and there is almost nothing to read. On the left sits a chart: four similar things side by side, each with three short feature bullets and a line giving a price and a validity or a term. Four transit passes. Four gym memberships. Four campsites. Four storage units. On the right sit eight questions.
The total word count in Part 2 is the lowest in the section, and candidates relax. Then eight minutes disappear and two questions are still unanswered. Understanding why that happens is most of what this lesson is for.
The eight questions come in two blocks. Questions 1 to 5 are blanks inside a short email in which one person walks a friend through the four options and recommends one. Questions 6 to 8 are sentence stems about that email's reasoning: why it ruled something out, what it assumed, what would change the recommendation.
It is a cross-referencing task, not a reading task
Here is the structural fact that decides your score. The email never contains the answer. It contains the person, their constraints and the shape of an argument, with the load-bearing facts cut out and replaced by drop-downs. Those facts sit in the chart, three columns to the left, in a bullet you have not read closely yet.
So Part 2 is not testing whether you can read a chart. It is testing whether you can hold two documents against each other and notice when they disagree. Candidates who answer from the email alone are answering from a text that has been deliberately hollowed out, and they will be right by accident sometimes and wrong reliably.
This is also where the time goes. Eight minutes for eight questions is about 60 seconds each, which sounds comfortable until you notice that each minute has to contain two journeys: one into the email for the constraint, one into the chart for the fact. The reading is not slow. The two trips inside the minute are.
The four options are near-identical on purpose
Look at any Part 2 chart and count what the four options share. Most of it. They will all have opening hours, they will all have a price, most will have some form of access or duration line. The similarity is deliberate.
Each option usually carries only two or three properties that no other one has. Those are the discriminators, and every single question in the part is built on one of them. Everything else on the chart is padding designed to cost you time you do not have.
The move that follows is simple and almost nobody makes it. Before you answer anything, ask what each option has that the other three do not. Two or three things each, twelve items or fewer in total. Any feature shared by three of the four columns is automatically not the reason one column wins or loses, so it can be ignored for the rest of the part.
Doing this takes about 45 seconds and it turns a chart of twelve bullets into a chart of eight that matter. That is the difference between a part you finish and a part that finishes you.
Spend your first 45 seconds finding what each of the four options has that the other three do not. Every question in the part is built on one of those discriminators, and anything shared by three columns can be ignored entirely.
How the wrong options are built
Part 2 distractors are the most mechanical in the section, and the most effective, because they are almost always true. The standard construction is a genuine feature copied from a different column of the same chart.
Consider a question about the weekend pass, with an option mentioning unlimited evening travel. Unlimited evening travel is real. It is printed on the chart. It belongs to the monthly pass. The option survives any check of the form did I read this somewhere? and fails only the check that matters: does this feature belong to the option this sentence is about?
The defence is a single habit. Before you open a drop-down, say out loud in your head which of the four options this sentence is about. Then read only that column. Almost every wrong answer in Part 2 falls away without any thought at all once the column is named first, because you are no longer choosing between four true things, you are checking one column.
A second, rarer family reverses a comparison. The chart shows option C cheaper than option D; the email blank offers a phrasing that makes D the cheaper one. These are easy to catch and easy to miss under time pressure, which is why the last blank in a rushed Part 2 is the one most often lost.
Name the column before you open any drop-down. Almost every wrong option in Part 2 is a true feature belonging to a different option, and naming the column first makes them disappear without analysis.
Let the email's grammar narrow the field
The email is written by a person making an argument, and the way they write constrains what can legally fill each blank before you look at the chart at all.
Phrases like the only one that, both of those, neither of the cheaper two and twelve dollars less than are doing real work. The only one that tells you the feature in the blank must be unique to one column, which eliminates anything shared. Both of those tells you the blank refers back to two options already named. A comparative tells you two specific numbers have to be checked and the rest ignored.
Read each sentence whole, including what comes after the blank, before you decide. The words on the far side of a drop-down are as constraining as the words before it, and candidates who read only up to the blank throw away half the evidence.
This is also how questions 6 to 8 are answered. Those three stems ask why the email reached the conclusions it did, which means the evidence is the argument itself: which constraint ruled out which option, and what would have to change for the recommendation to change.
A procedure for the eight minutes
Part 2 rewards a fixed procedure more than any other part, because the reasoning is mechanical once you are pointed at the right two lines.
Forty-five seconds on the shape of the chart: what the four things are, which dimensions are compared, and what each column uniquely has. Do not study the bullets before this, because you do not yet know what you are looking for.
Sixty seconds on the email, read straight through with the blanks empty. What you are mining for is constraints: what this person cannot have. A budget. A date. A vehicle that will not fit. A commute of a particular length. Every blank is a collision between one constraint and one chart bullet, so the constraints are the key that opens the chart.
Then about five minutes on the eight questions, naming the column before each one, which leaves just over a minute of slack for the blank that fights you. If a blank has taken more than 90 seconds, choose the option that matches the named column best and move. There is no penalty for a wrong answer, and the two questions after it are worth the same as the one in front of you.
Read each email sentence past the blank to its full stop. Words like "the only one that" and "twelve dollars less" sit after the drop-down as often as before it, and they often narrow four options to one on their own.
Practise it
Five questions on the diagram part: where the answers live, how the distractors are built, and how to spend the eight minutes.