9.4 The Two Official Critical-Thinking Item Types: Data Propositions & Deducibility

9.4 The Two Official Critical-Thinking Item Types: Data Propositions & Deducibility

Key Exam Fact: Bocconi describes the critical-thinking area — 9 of the 50 questions, 18% of the test — in terms of exactly two item types. In the first, "data and statements will be presented. The applicant will have to choose which statements are true (or false) depending only on the data provided." In the second, "a short passage will be presented that includes a number of assertions intended to convey information or convince the reader of an opinion. The passage is followed by an assertion, and the applicant must decide whether this information, based on what is given in the passage, is true, false, or it cannot be deduced from the text because information is missing." Everything else in this chapter is preparation for these two items.

Bocconi's own definition of critical thinking is "the ability to actively and skilfully conceptualize, apply, analyze, summarize and/or assess information." In practice, the University tests that ability through a narrow and highly regular pair of formats. This section drills both.

There is a scoring wrinkle worth knowing before you start. Bocconi's penalty is −0.20 for a wrong answer, but −0.33 "for questions of the critical thinking area that give only three possible answer options." A true / false / cannot-be-deduced item has exactly three options. These items are therefore the most expensive questions on the test to get wrong — and, as shown below, the ones where careless reading is easiest.

---

1. Item Type One: Data Propositions

You are given a compact body of data — a small table, a set of figures, a short list of factual statements — followed by candidate statements. Your task is to decide which statements are true, or which are false, using the supplied data and nothing else.

The Closed-Data Principle

The four words that decide these items are Bocconi's own: "depending only on the data provided." Three consequences follow, and every one of them is regularly violated by candidates who know the subject matter well:

1. Outside knowledge is inadmissible. If the data says a firm's revenue rose while its margin fell, you may conclude what follows arithmetically. You may not conclude anything that requires knowing how firms in that industry usually behave.
2. Plausible is not the same as supported. A statement can be overwhelmingly likely in the real world and still be unsupported by the data in front of you.
3. Unsupported is not the same as false. If a statement asks about something the data simply does not address, it is not established — but neither is its opposite. Which verdict that maps onto depends on how the question is phrased, which is why you read the stem before the statements.

The Verification Procedure

1. Read the stem first. Are you asked which statements are true, or which are false? The two produce opposite answers from the same analysis, and swapping them is the most common careless error on this item type.
2. Inventory the data. Note the units, the time periods, the totals, and — critically — what is absent. A table showing revenue by division for one year says nothing about growth.
3. Take one statement at a time. Locate the specific figures or clauses that bear on it. If you cannot point at the evidence, the statement is not established.
4. Interrogate the quantifier and the scope. "All", "only", "every", "at least", "more than", "the largest" are load-bearing. A statement that is true for two of three divisions is false if it claims all three.
5. Watch for derived quantities. Many statements are true only after a one-step calculation — a share, a difference, a ratio. Do the arithmetic; do not eyeball it.

Worked Example: A Data-Proposition Item

Data. A logistics group reports the following for last year. Northern division: revenue €48m, operating cost €40m. Central division: revenue €30m, operating cost €27m. Southern division: revenue €22m, operating cost €20m. The group opened no new divisions during the year.

Statements. (i) The Northern division was the most profitable in absolute terms. (ii) Every division was profitable. (iii) The Southern division had the highest operating margin. (iv) Group revenue grew during the year.

Task. Which statements are true on the basis of the data alone?

• Statement (i). Operating profit is revenue minus operating cost: Northern 48 − 40 = €8m; Central 30 − 27 = €3m; Southern 22 − 20 = €2m. Northern is the largest in absolute terms. TRUE.
• Statement (ii). All three profits are positive. TRUE.
• Statement (iii). Margin is profit divided by revenue: Northern 8/48 ≈ 16.7%; Central 3/30 = 10.0%; Southern 2/22 ≈ 9.1%. Southern is the lowest, not the highest. FALSE. Note the trap: Southern is the smallest division, and "smallest" invites a careless reader to accept any superlative attached to it.
• Statement (iv). The data gives one year only. Growth requires a prior-year comparison the data does not contain. The sentence about new divisions is a decoy: it constrains the structure of the group, not its revenue trend. NOT ESTABLISHED — and therefore not among the true statements.

Answer: (i) and (ii).

---

2. Item Type Two: True / False / Cannot Be Deduced

A short passage makes a case. One assertion follows. You return exactly one of three verdicts.

The Three Verdicts, Defined Precisely

Bocconi's own wording for the third verdict is the one to memorise: it "cannot be deduced from the text because information is missing."

The Decision Algorithm

1. Find the evidence. Locate the part of the passage the assertion is about. No relevant text at all is a strong signal for cannot be deduced.
2. Test for contradiction first. Ask whether the passage and the assertion can both be true at once. If they cannot, the verdict is false and you are finished.
3. Then test for entailment. Ask whether the passage forces the assertion. Not "supports", not "suggests", not "makes likely" — forces. If yes, the verdict is true.
4. Everything else is "cannot be deduced." This includes assertions that are probably right, assertions the author would almost certainly agree with, and assertions that hold for most but not all of the cases the passage covers.

The Distinction That Decides These Items

The single most valuable habit is refusing to collapse false into cannot be deduced or the reverse.

• Passage: "The reform reduced processing times in the three northern regions."
• Assertion A: "The reform reduced processing times in every region." → Cannot be deduced. The passage is silent on the other regions; it does not claim they were unaffected, and it does not claim they were.
• Assertion B: "The reform had no effect on processing times in the northern regions." → False. This collides directly with the passage.
• Assertion C: "Processing times fell in at least one northern region." → True. Three regions saw a reduction, so at least one did.

Notice how little the verdicts depend on subject knowledge and how much they depend on quantifiers: every, no, at least one.

Worked Example: A Passage-Plus-Assertion Item

Passage. Since the introduction of the digital filing portal, the national business registry has cut the average time to incorporate a company from eleven days to four. The registry attributes the improvement to automated identity verification, which removed the manual document review that had been the slowest step. Registry officials note that the portal has not reduced the incorporation fee, and that the fastest filings are concentrated among applicants who already hold a verified digital identity.

Assertion 1: "Incorporating a company through the registry is now cheaper than it was before the portal was introduced." Verdict: FALSE. The passage states that the portal "has not reduced the incorporation fee." The assertion collides with an explicit clause; that is a contradiction, not a silence.

Assertion 2: "Manual document review was the slowest step in the incorporation process before the portal was introduced." Verdict: TRUE. The passage says automated identity verification "removed the manual document review that had been the slowest step." The assertion restates an explicit claim.

Assertion 3: "Applicants without a verified digital identity now incorporate a company in fewer than eleven days." Verdict: CANNOT BE DEDUCED. The passage gives an overall average of four days and says the fastest filings cluster among applicants who already hold a verified digital identity. It gives no figure for the other group. An average of four days is consistent with that group still taking eleven days or longer, and equally consistent with them taking six. Information is missing — which is exactly Bocconi's definition of the third verdict. The temptation here is real: the assertion is very probably true in the world. That is not the standard.

---

3. The Trap Catalogue

---

4. Timing and Scoring on These Items

These items are short to read and fast to decide once the method is automatic — typically 45 to 70 seconds — which makes them one of the few places on a 90-second-average test where you can bank time for a multi-step geometry item sharing the same screen.

The scoring asymmetry deserves one last look. On a three-option item, a blind guess has an expected value of about +0.11 points, because a one-in-three chance at +1.00 is set against a two-in-three chance at −0.33. Eliminate one verdict — usually the easiest move, since false can often be confirmed or dismissed by locating a single clause — and the two remaining verdicts give an expected value of roughly +0.335. Since the Online Bocconi Test is forward-only, that elimination has to happen before you click Next, not in a review phase that does not exist.