3.10 Decision Making & Data Analysis in Reasoning

Part 1: Decision Making

A decision-making item gives a set of eligibility conditions, one or more exception clauses, and then several candidate cases. For each case you choose one of five standard responses: the candidate is selected; is not selected; the case is referred to a higher authority; data are inadequate; or some other stated routing.

The condition-matrix method

This is the only reliable approach under time pressure.

1. List the conditions vertically as C1, C2, C3 and so on, before reading any case.
2. List the exceptions separately, noting exactly which condition each exception relaxes and where it routes the case.
3. For each candidate, tick or cross every condition in order.
4. If all conditions are satisfied, the candidate qualifies.
5. If exactly one condition fails, check whether an exception covers that specific condition. If it does, apply the routing the exception specifies. If it does not, the candidate is rejected.
6. If two or more conditions fail, the candidate is rejected regardless of exceptions, unless the question says otherwise.
7. If any condition cannot be evaluated because the case omits the relevant fact, the answer is data inadequate.

Worked example

Conditions for selection as Management Trainee (Mechanical): C1. Holds a B.E. or B.Tech in Mechanical Engineering with at least 60% marks. C2. Is not more than 30 years of age as on the cut-off date. C3. Has secured the qualifying marks in both papers of the CBT.

Exceptions: E1. If C2 fails but the candidate belongs to a category entitled to age relaxation, refer the case to the Recruitment Head. E2. If C1 fails only on the percentage but the candidate holds a postgraduate degree in the same discipline, refer to the General Manager (HR).

Case: Anand holds a B.Tech in Mechanical Engineering with 58% marks and an M.Tech in Mechanical Engineering. He is 27 years old and has qualified in both papers.

Work the matrix: C1 fails on percentage; C2 passes; C3 passes. Exactly one condition fails, and E2 addresses precisely that failure while Anand holds the required postgraduate degree. The correct response is therefore refer to the General Manager (HR) — not rejection.

This is the archetypal trap. Candidates who stop at "58% is below 60%" mark rejection and lose the mark.

Discipline points

• Read all conditions before any case. Reading case-first causes conditions to be missed.
• Note whether conditions are joined by and or or. An or condition needs only one branch satisfied.
• Watch cut-off dates and the phrase as on, which fixes the reference date.
• Never import a real-world assumption. If the case does not state a category, you may not assume one.

Part 2: Data Analysis in Reasoning

The syllabus lists Data Analysis - Charts, Tables, Graphs under Logical Reasoning as well as listing data interpretation arithmetic under Quantitative Aptitude. The reasoning version differs in emphasis: it asks about ranking, comparison, consistency and sufficiency rather than about computing an exact figure. Detailed arithmetic technique for percentages, averages and ratios in charts is covered in the Numerical Ability chapter; this section covers the reasoning layer.

Reading a table efficiently

Before answering anything, spend fifteen seconds on the structure:

1. What do the rows represent, and what do the columns represent?
2. What are the units? Tonnes, lakh tonnes and million tonnes are not interchangeable, and a mismatch is a deliberate trap.
3. Is any row a total or an average? Totals must not be double counted in a sum.
4. Are any cells blank or marked as not available? These usually make a related question data-inadequate.

Comparison without computation

Most reasoning-flavoured questions can be answered by comparison rather than calculation.

The single most exploited confusion is between absolute increase and percentage increase. An entity rising from 10 to 15 has an absolute increase of 5 and a percentage increase of 50%; one rising from 100 to 120 has a larger absolute increase of 20 but a smaller percentage increase of 20%. Read which one the question asks for.

Worked example

Coal despatch by four subsidiaries, in million tonnes:

Subsidiary	Year 1	Year 2
P	120	138
Q	80	96
R	150	165
S	60	75

Which subsidiary recorded the highest percentage growth?

Compute growth ratios rather than differences: P grows by 18 on 120, which is 15%; Q by 16 on 80, which is 20%; R by 15 on 150, which is 10%; S by 15 on 60, which is 25%. The answer is S, even though R had the largest absolute despatch and P the largest absolute increase.

Which had the largest absolute increase? That is P, with 18. The two questions have different answers from the same table, which is precisely why the distinction matters.

Data sufficiency

A data-sufficiency item gives a question and two statements, asking whether the statements suffice. The standard procedure:

1. Evaluate statement I alone. Is the question answered?
2. Evaluate statement II alone, deliberately forgetting statement I.
3. Only if neither alone suffices, evaluate them together.
4. Choose among: I alone is sufficient; II alone is sufficient; either alone is sufficient; both together are needed; both together are still insufficient.

The discipline that matters is step 2 — candidates routinely carry information from statement I into their assessment of statement II and wrongly conclude that II alone suffices. Cover statement I physically on the screen if that helps.

Note also that you are asked only whether the question can be answered, not what the answer is. Recognising that a unique value is determined is enough; computing it wastes time.