The role of DILR in CAT
Data Interpretation and Logical Reasoning is the second CAT section: 22 questions in 40 minutes in CAT 2025. It is organised in sets, each built around a scenario, table, chart or set of conditions followed by several questions. Candidates often find it the most volatile section, because a set that clicks gives you four or five questions in ten minutes, and a set that does not can leave you with nothing after fifteen. Its volatility is why many candidates see their overall percentile move most with DILR, and why practising the choice of sets is as important as practising the sets themselves.
What the section contains
| Set type | What it looks like | What it tests |
|---|---|---|
| Arrangements and selections | People or objects placed in positions or groups under conditions | Systematic deduction; case analysis |
| Games and tournaments | Results of matches, points tables, rankings with missing data | Constraint handling; working with partial information |
| Tables, bar charts and line graphs | Data on sales, population, production or similar, with percentage and ratio questions | Calculation speed; reading data |
| Caselets | A text-based scenario with many numbers | Extracting structure from text |
| Networks and routes | Nodes and links with distances, costs or capacities | Path logic; optimisation |
| Venn-style and set logic | Overlapping groups with membership conditions | Counting; consistency checks |
Which formats appear, and how many sets there are (usually four or five, with four to six questions each), changes from year to year. Questions may be MCQs or type-in questions. The IIMs do not publish a syllabus, so prepare across all of these families rather than guess which will appear.
Why DILR is a selection game
A set has an all-or-nothing character: the questions within it build on the same structure, so once you have cracked the structure, most questions become quick. If you have not, none is accessible. That means the question to ask in the first minutes is not "Can I solve this question?" but "Can I crack this set?". The sets in a paper are not equal: typically one or two are approachable, one or two are moderately hard and one is much harder than the rest. If you commit to the hardest set, you may spend 15 minutes and score nothing.
A triage method for the first six minutes
- Read all the sets quickly (about 60 to 90 seconds each), not trying to solve, only to understand the structure.
- Look for a way in. A set is promising if some information is definite (a fixed position, a known total), the conditions are few and clear and at least one question can be answered after one or two deductions.
- Warning signs: many variables, several "either/or" conditions, a long caselet with many interdependent numbers or conditions that only make sense together.
- Rank the sets from easiest to hardest and write the order on your rough sheet.
- Commit to the easiest. Set a mental limit of about eight to ten minutes before you reconsider.
Triage costs time, but a set chosen well returns many times its cost. In mocks, experiment with the length of triage (three minutes versus six) and see which produces more correct answers.
Solving a set: habits that prevent errors
- Draw a clean diagram or table first. Most errors come from a messy first diagram. Give it a third of your set time.
- Write conditions in short form next to the diagram, and tick each as you use it.
- Use cases when forced to. When a condition gives two possibilities, label them A and B and carry both until a later condition removes one. Do not choose one and hope.
- Verify the final arrangement against every condition before you start on the questions.
- Re-read each question's exact ask, especially questions that add a new condition ("if X moves to position 3...").
Calculation in data sets
In table and chart sets, the difficulty is not the maths but the volume of calculation. Practise approximation, comparison of fractions by cross-multiplication and the use of ratios instead of absolute values. Our data interpretation practice and percentage drills build the fluency, and how to practise data interpretation faster explains the habits in detail.
A worked example: solving a small arrangement set
Here is a short original set, much easier than a real CAT set, to show the method. Five friends, A, B, C, D and E, sit in a row of five chairs numbered 1 to 5 from left to right. The conditions are: (1) B sits immediately to the left of C; (2) D sits in chair 5; (3) A sits somewhere to the left of B; (4) E does not sit next to A.
Step 1: fix what is certain. D is in chair 5, so A, B, C and E share chairs 1 to 4.
Step 2: place the linked pair. B and C occupy two adjacent chairs with B on the left. The possible positions for (B, C) are (1, 2), (2, 3) and (3, 4).
Step 3: test each case against the other conditions.
- (B, C) = (1, 2): A must be to the left of B, but nothing is left of chair 1. Rejected.
- (B, C) = (2, 3): A must be in chair 1, leaving chair 4 for E. A is in 1 and E in 4, so they are not adjacent. Valid: A, B, C, E, D.
- (B, C) = (3, 4): chairs 1 and 2 remain for A and E, and they are adjacent whichever way round they sit, so condition (4) fails. Rejected.
Step 4: verify and answer. The only arrangement is A B C E D. So C sits in the middle, three people sit between A and D, and E sits immediately to the left of D. If condition (4) were dropped, the case (3, 4) would give two more arrangements, so there would be three in all. Questions of the form "if this condition is dropped" or "if X moves to chair 3, which of the following must be true?" are common in real sets, and having your cases laid out is what makes them fast.
Three habits from this example matter more than the puzzle itself: fix certain things first, list the cases for the linked pair and test each case against all remaining conditions before concluding. A real CAT set has more people, more attributes and subtler conditions, but the structure of the solution is the same.
A worked example: a data set and the calculations it invites
Sales (in units) of three products across four quarters:
| Product | Q1 | Q2 | Q3 | Q4 | Total |
|---|---|---|---|---|---|
| P1 | 120 | 150 | 180 | 210 | 660 |
| P2 | 90 | 110 | 100 | 140 | 440 |
| P3 | 60 | 80 | 120 | 100 | 360 |
| All | 270 | 340 | 400 | 450 | 1,460 |
Typical questions and the fast way to answer them: What was the percentage growth in P1's sales from Q1 to Q4? The increase is 90 on a base of 120, which is 75%. In which quarter were total sales highest? Q4 (450). What share of total sales did P1 account for? 660 out of 1,460 is about 45%: you can see it quickly by rounding 1,460 to 1,500 and noting that 660 is a little under 45% of 1,500, then adjusting upwards slightly because the true denominator is smaller. If the options were 40%, 45%, 50% and 55%, you would not need to calculate precisely at all.
The lesson is the same one as in the arrangement example: understand the structure first, then choose the cheapest calculation that separates the options. Practise this with our data interpretation sets.
Time management in the 40 minutes
| Minutes | What to do |
|---|---|
| 0-6 | Triage: read every set, rank them |
| 6-17 | First set: aim for all questions |
| 17-28 | Second set |
| 28-37 | Third set, or the easiest questions of a fourth |
| 37-40 | Check type-in answers; mark anything left that you can answer quickly |
This is an example of the shape, not a rule. If your first set takes only six minutes, you have time for a fourth. If it takes sixteen, abandon the idea of a fourth and focus on accuracy in the third. The decision to abandon a set is the hardest one in the section, so practise it in mocks, and see how to pick which questions to skip.
Reading signals: which sets tend to pay off
| Signal in the set | What it usually means | Action |
|---|---|---|
| Some placements or totals are given outright | There is a definite entry point | Good candidate; start there |
| Conditions are mostly "A is next to B" type with few unknowns | Small case tree | Good candidate |
| Many "if-then" statements chained together | Deep case analysis | Try only if other sets fail |
| A long caselet with several interdependent quantities | High risk of calculation errors | Attempt later, after safer sets |
| Table with clean numbers and direct questions | Calculation-heavy but low risk | Often a good first set |
These signals are rules of thumb, not guarantees. Treat them as a way to make a quick first judgement and then correct it from experience: after each mock, compare your prediction of a set's difficulty with how the set actually went. Over ten mocks you will have a much better feel than any rule gives.
Negative marking in DILR
A wrong MCQ costs one mark; type-in questions carry no penalty. Because questions within a set are linked, a wrong deduction early in a set can make several answers wrong together, which makes errors in this section costly. Verify your structure before answering and be cautious about guessing on MCQs from a set you have not cracked.
How to practise: a progression
- Untimed sets (weeks 1-3): solve sets from each family without a clock. The goal is to learn the method and the typical patterns.
- Timed single sets (weeks 3-6): give yourself 10 to 12 minutes per set and stop at the limit. Review what you would have done differently.
- Triage practice (weeks 6-10): take four-set sectionals with a 40-minute clock and a strict triage step.
- Full mocks (weeks 10 onwards): see how DILR behaves in the middle of a two-hour paper.
Always redo the sets you skipped after the clock, because that is how you learn which signals predict a solvable set.
Analysing your DILR result
| Pattern in your mock | Probable cause | Fix |
|---|---|---|
| Four sets touched, few questions correct | Poor triage or low persistence | Commit to three sets; set a ten-minute limit |
| A set started, abandoned, then the time was gone | Sunk-cost thinking | Decide the abandonment point before you start the set |
| Correct structure, wrong answers | Calculation or reading errors | Verify the structure; re-read each question's ask |
| Skipped sets turn out to be easy afterwards | Triage too cautious | Practise finding the first deduction faster |
| Skipped sets were truly hard | Triage worked | Keep the method; work on speed on the sets you did attempt |
Common mistakes in DILR preparation
- Practising only sets you can solve, which hides your weak families.
- Reading the solution too early. Struggling for ten minutes before you look at the solution is part of the learning.
- Not timing sets until late in preparation.
- Using scratch paper carelessly; a clean layout of the diagram saves minutes.
- Assuming that more sets attempted means a better score.
Related pages
See the CAT hub, the exam pattern and the mock test strategy. The other sections are VARC and Quantitative Aptitude. For similar reasoning topics, see syllogisms and blood relations, and for a worked review, how to use mock tests for CAT preparation.
Note: the DILR description above reflects CAT 2025 (22 questions in four or five sets); the number and style of sets differ each year, so rely on the official notification and recent papers for the latest picture.
