Logical reasoning in placement tests: puzzles, seating and syllogisms
The reasoning section of a fresher placement test checks whether you can follow rules and arrange information quickly. The most common question types are series, coding-decoding, blood relations, direction sense, syllogisms, seating arrangement and small puzzles. None of them needs advanced maths, so a few weeks of steady practice can move this section from your weakest to your safest.
Keep one caution in mind: companies do not publish a topic-wise weightage for reasoning. The counts below come from official pages where available, and otherwise from past papers and test-taker reports, which are estimates and can change from drive to drive.
Quick facts
| Item | What we found |
|---|---|
| TCS NQT | Official: Reasoning Ability is a 25-minute part of the 75-minute Foundation section |
| TCS sample paper | Official sample: 11 reasoning questions out of 51, no negative marking |
| Infosys | Reported: about 15 reasoning questions in 25 minutes (estimate, not official) |
| Level | School-level logic, asked under time pressure |
| Target speed | About 1 to 1.5 minutes per question in practice |
| Official topic weightage | Not published by the companies |
What the official TCS material shows
TCS is one of the few recruiters that publishes its test structure. Its All India NQT page lists three 25-minute parts in the Foundation section: Numerical Ability, Verbal Ability and Reasoning Ability. The Advanced section adds a part on advanced quantitative and reasoning ability, which matters for the higher-paying Digital and Prime roles.
TCS has also published a sample paper for its Smart Hiring and B.Sc Ignite tests. Its reasoning section is a good picture of the variety you should expect:
- letter series and number series
- analogies (letters, numbers and figures)
- a syllogism (statements and conclusions)
- a room-allocation puzzle
- applying eligibility rules to a person
- courses of action from a statement
- mirror images
Here is the number series from that paper: 78, 156, 468, 1872, ? The multipliers are ×2, ×3 and ×4, so the next is ×5, and the answer is 9,360.
The question types, in the order we would study them
Based on past papers, the topics group roughly as below. These priorities are estimates, not official figures.
| Priority | Topics | Why |
|---|---|---|
| First | Number and letter series, coding-decoding, analogies | Quick to learn and asked almost everywhere |
| First | Blood relations, direction sense | Rule-based; a simple diagram solves most of them |
| Second | Syllogisms, statements and conclusions | One method (Venn diagrams) covers nearly all questions |
| Second | Seating arrangement and ranking | Take longer, but sets of 3 to 5 questions share one diagram |
| Third | Puzzles (scheduling, allocation), data sufficiency | Most time-consuming; attempt once the basics are fast |
| Last | Figure-based (mirror images, odd one out) | Usually only one or two questions |
Methods that save the most time
Syllogisms: draw the smallest possible Venn diagram. For "Some hill is land, all land is plateau, a few lands are peaks", draw land inside plateau, then let hill and peak each overlap land. Now check whether a conclusion such as "Some peaks are hills" must be true. Since you can draw peak and hill without touching each other, it is not certain. In the TCS sample paper, the answer to this question is that neither conclusion follows. The rule: a conclusion follows only if it is true in every diagram you can draw.
Blood relations: use symbols. Write men as squares and women as circles, use a horizontal line for a married couple and a vertical line for parent and child. Long sentences such as "the only son of my grandfather's daughter" become a two-second diagram.
Seating arrangement: fix the definite clues first. Start with clues that place a person exactly ("A sits third from the left"), then add clues that link two people ("B sits next to A"), and keep "not" clues for last. For circular tables, remember that left and right depend on whether people face the centre or face outward.
Puzzles: make a grid before reading the second clue. Put people on one side and days, floors or items on the other, and mark ticks and crosses. Most mistakes in puzzles come from holding the information in your head.
Know when to skip. If a seating or puzzle set is not opening up after about 3 minutes, move on and come back at the end.
A two-week plan
- Days 1 to 4: number series, letter series, coding-decoding and analogies. Do 25 questions per topic and read every solution.
- Days 5 to 7: blood relations, direction sense and ranking and order. Draw a diagram for every question, even the easy ones.
- Days 8 to 10: syllogisms, inequalities and odd one out.
- Days 11 to 14: seating arrangement, plus clocks and calendars. End each day with 15 mixed questions in 20 minutes.
After that, add a timed reasoning section to every full mock. All the topics above have free questions with solutions on Target Campus placement practice. For the quant side of the same tests, see our guide to quant topics in placement tests.
Frequently asked questions
Is logical reasoning harder than quant in placement tests?
For most students it is easier to improve, because it depends on methods rather than formulas. Seating and puzzle sets are the exception: they take time, so practise them under a timer.
Is there negative marking in the reasoning section?
It depends on the company and the drive. The official TCS sample paper has no negative marking, and test-takers report none for Infosys, but always read the instructions on the test screen before you start guessing.
Do I need to practise figure-based reasoning?
A little. The TCS sample paper includes mirror images and figure analogies, but they are usually only one or two questions. Learn the basic types and spend more time on series, arrangements and syllogisms.
How many questions should I practise?
Aim for about 300 to 400 questions across the topics over two weeks, with every wrong answer noted in a mistake notebook. Quality matters more than volume: understand why each wrong option is wrong.
