Mass-recruiter online tests: typical sections and a 4-week drill
In this article (6 sections)
Related programme: the Data Analytics with Generative AI course (3–4 months; ₹90,000 (inclusive of GST) · EMI from ₹4,000/month).
Large-volume fresher hiring tests usually combine some of these sections: quantitative aptitude, logical reasoning, verbal ability, a technical or computer-fundamentals section, and one or two coding problems, sometimes followed by a short written task. Each company publishes its own pattern in the test invitation, and patterns change between years, so prepare for the section types rather than a remembered question count. The four-week drill below covers all of them and ends with a remote-proctoring check.
What each section usually checks
| Section | What it tests | How it is usually scored | Your best preparation |
|---|---|---|---|
| Quantitative aptitude | Arithmetic patterns under time pressure | Correct answers, sometimes with negative marking | Timed topic blocks, then mixed sets |
| Logical reasoning | Series, arrangements, syllogisms, data sufficiency | Same as above | Learn the first move for each pattern |
| Verbal ability | Grammar, vocabulary in context, comprehension | Same as above | Rule list plus daily reading |
| Technical MCQs | Programming basics, operating systems, networks, databases | Correct answers | One revision pass of each fundamentals subject |
| Output prediction / pseudocode | Tracing code | Correct answers | Trace tables on paper |
| Coding | One or two problems in an online editor | Hidden test cases passed | Timed practice in a similar editor |
| Written task | Email or short essay | Rubric or automated scoring | Plain, structured writing |
Some tests have sectional cut-offs, meaning a weak section can end your attempt even if your total is high. Others use an overall score. The invitation or instructions page usually tells you whether sections are timed separately and whether you can move back to earlier questions. Read it the day before, not during the test.
Negative marking changes the guessing rule
Do the arithmetic once so you never have to think about it in the test. Take a question with four options.
Scheme A: +1 correct, −0.25 wrong. Blind guess: expected score = (1/4 × 1) − (3/4 × 0.25) = 0.25 − 0.1875 = +0.0625. After eliminating one option: (1/3 × 1) − (2/3 × 0.25) = 0.333 − 0.167 = +0.167. So guessing is mildly positive, and much better after elimination.
Scheme B: +1 correct, −1/3 wrong. Blind guess: (1/4 × 1) − (3/4 × 1/3) = 0.25 − 0.25 = 0. After eliminating one option: (1/3 × 1) − (2/3 × 1/3) = 0.333 − 0.222 = +0.111. So never guess blind, but do guess once you can rule something out.
No negative marking: answer everything before time runs out.
These are expected values over many questions, not a promise for one question — but a test has many questions, which is exactly when expected value is the right guide.
The four-week drill
About 90 minutes on weekdays and three hours on one weekend day.
| Week | Weekdays | Weekend |
|---|---|---|
| 1 | Quant topics, one per day, 20 questions; read one fundamentals chapter | 60-minute mixed aptitude set; review errors |
| 2 | Logical and verbal topics, one per day; trace five code snippets daily | Full-length mock with all sections |
| 3 | Mixed timed sets of 20; two coding problems a day in an online editor | Full mock; compare section scores with week 2 |
| 4 | Revise error log only; one coding problem and one written task daily | Two full mocks on different days, same time as your real test |
In week 4 stop learning new topics. Your score now depends on speed and accuracy on what you already know.
Worked practice for the sections: quantitative aptitude, logical reasoning and output-prediction questions.
Track your mock scores properly
Record each mock in a sheet with one row per section: attempted, correct, wrong, time used, and accuracy (correct ÷ attempted). Accuracy tells you more than the raw score. If accuracy is below about 70% in a section, you are guessing too much or rushing; slow down. If accuracy is high but attempts are low, you are too slow; practise recognition drills.
Remote-proctored test checklist
Many first rounds are taken from home under webcam proctoring. Problems here end attempts that knowledge would have passed.
- Run the system-compatibility check the invitation links to, on the same laptop and network you will use, at least a day early.
- Keep the original ID proof named in the instructions next to you.
- Charge the laptop and keep it plugged in; close other applications, messaging apps and browser tabs.
- Use a quiet room with light in front of you, not behind. Tell family the exact time window.
- Keep a mobile hotspot ready in case your Wi-Fi drops, but check whether the instructions allow switching networks.
- Do not keep papers or a second screen on the desk unless rough sheets are explicitly permitted.
- Note the support email or helpline from the invitation before you start.
After the test
Write down, from memory, which section felt weakest and why. Results from large drives can take weeks, and you may sit several similar tests in one season; the note is what lets the next attempt improve. Be careful with anyone who contacts you afterwards asking for a fee to "clear" your result or release an offer — legitimate hiring does not work that way.
The same timed accuracy you build here is what an analyst needs when a report is due at 10 a.m. If you want to build that habit on real SQL and spreadsheet work, the Data Analyst programme sets weekly timed deliverables.
Frequently asked questions
Should I guess in an online test with negative marking?
Work it out from the marking scheme in the instructions. With four options, +1 for correct and −0.25 for wrong, a blind guess has a small positive expected score, and eliminating even one option improves it. With −1/3 per wrong answer, a blind guess has zero expected value, so guess only after eliminating an option.
Where can I find the exact pattern of a company's test?
Use the company's own careers page, the official test invitation email and your placement cell's briefing. Pattern details shared on forums are often outdated or mixed up between years and roles, so treat them as rough practice only.
Pankit Kumar has 10 years in Data Science & AI, building and shipping production systems in regulated pharma and clinical environments. He is a freelance trainer at Boston Institute of Analytics, AnalytixLabs and Scaler, and has taught this material to thousands of working professionals.
This article is part of our Data Analytics with Generative AI programme — 3–4 months. The full analyst stack — Excel, SQL, Power BI and Python pipelines — then a generative-AI layer you can prove is right.
Explore Data Analytics with Generative AI