How placement aptitude tests work

AMCAT’s Logical Ability module gives you 12 questions in 15 minutes, 75 seconds each, while SSC CGL’s Tier-I allows 36 seconds a question. A wrong answer costs nothing in TCS iON NQT, but a third of the question’s marks in a GATE multiple-choice question.

A placement aptitude test is a timed online test of working with numbers, logic and language that employers use to screen freshers in large numbers. Nearly every one covers the same three areas: quantitative ability (arithmetic, number work and reading data from tables and charts), logical reasoning (patterns, arrangements and deduction) and verbal ability (grammar, vocabulary and reading). What differs from test to test is the rulebook: how the time is given, what a wrong answer costs, how the score is made and what comes attached. Those rules change how you should practise more than the content does, and each test publishes them on its own pages.

Every test covers the same three areas, under its own rules

Each test names its sections in its own words, but the content overlaps almost completely. This is what their own pages list:

Two things stand out. The shared core is large: percentages, ratios, series, arrangements, grammar and reading appear under different headings in all five, which is why this track is organised by topic rather than by test. And two kinds of material reach beyond that core. Spatial work, shapes that are moved, rotated, scaled, mirrored, folded and cut, is GATE’s spatial aptitude and part of the NQT’s visual-spatial reasoning, and it has a module of its own in this track. AMCAT adds a personality module and optional domain modules, among them Automata Fix, in which you fix or complete code.

Before you plan your preparation for a test, answer four questions from its own page:

Each answer becomes a practice decision. Step through the four questions on two tests whose rules differ on every point, AMCAT and the eLitmus pH Test:

The same four questions gave two plans: paced practice module by module for AMCAT, slower and deeper practice without blind guessing for eLitmus. The next four sections answer the questions for all five tests, and add one recruitment test, SSC CGL, for contrast.

Seconds per question is the pace to practise at

A test gives its time in one of two ways: one clock for the whole paper, or a fixed clock for each section. With one clock you can move minutes from a section you find easy to one you find hard; with a sectional time limitsectional time limit: A fixed amount of time for one section of a test. Minutes left over when that section ends cannot be spent on another section. you cannot. CAT 2026 is the strict case: its press release gives exactly 40 minutes to each of the three sections and does not let you switch from one section to another while a section runs. SSC CGL’s Tier-I runs a 15-minute timer for each of its four subjects. AMCAT’s syllabus gives each module its own duration, eLitmus says its pH Test is designed to be completed in two hours, and a GATE paper lasts three hours for 65 questions.

Seconds per question is the number to train with, because it is what your timed practice should feel like. The program divides each published time by its question count. Press Edit, add a row for the test you are taking, and run it again:

// Seconds per question for each timed part of five tests, from the question counts and minutes that each test's own
// page publishes (the pages are linked under References). Edit a row, or add the test you are taking, and press Run.
const parts = [
  // [test and part, questions, minutes]
  ['AMCAT English', 18, 18],
  ['AMCAT Quantitative (Tech)', 16, 20],
  ['AMCAT Quantitative', 14, 20],
  ['AMCAT Logical Ability', 12, 15],
  ['AMCAT Information Gathering', 12, 15],
  ['AMCAT AMPI personality', 90, 20],
  ['eLitmus pH Test, whole test', 60, 120],
  ['GATE 2027, whole paper', 65, 180],
  ['SSC CGL Tier-I, each subject', 25, 15],
];

/** Seconds per question, with one decimal unless it is a whole number (the exact value is rounded once). */
function secondsEach(questions, minutes) {
  const tenths = Math.round((minutes * 60 * 10) / questions);
  return tenths % 10 === 0 ? String(tenths / 10) : (tenths / 10).toFixed(1);
}

for (const [name, questions, minutes] of parts) {
  const pace = `${questions} in ${minutes} min`;
  console.log(`${name.padEnd(28)} ${pace.padStart(13)} ${secondsEach(questions, minutes).padStart(6)} s`);
}

// What a slower pace does to a timed section: the questions you reach before its time runs out.
const [, questions, minutes] = parts[3];
for (const seconds of [75, 90, 120]) {
  const reached = Math.min(questions, Math.floor((minutes * 60) / seconds));
  console.log(`Logical Ability at ${seconds} s each: you reach ${reached} of ${questions}`);
}

AMCAT English                 18 in 18 min     60 s
AMCAT Quantitative (Tech)     16 in 20 min     75 s
AMCAT Quantitative            14 in 20 min   85.7 s
AMCAT Logical Ability         12 in 15 min     75 s
AMCAT Information Gathering   12 in 15 min     75 s
AMCAT AMPI personality        90 in 20 min   13.3 s
eLitmus pH Test, whole test  60 in 120 min    120 s
GATE 2027, whole paper       65 in 180 min  166.2 s
SSC CGL Tier-I, each subject  25 in 15 min     36 s
Logical Ability at 75 s each: you reach 12 of 12
Logical Ability at 90 s each: you reach 10 of 12
Logical Ability at 120 s each: you reach 7 of 12

What the numbers say:

• AMCAT’s English module gives one minute a question, and that minute covers its reading comprehension questions too, so a passage has to be read once, with the question in mind. Its Quantitative Ability (Tech) and Logical Ability modules give 75 seconds, and the second version of Quantitative Ability gives about 86.
• The AMPI personality module gives about 13 seconds a statement, which leaves time for a first reaction, not for deliberation.
• SSC CGL Tier-I is the fastest at 36 seconds a question, while the eLitmus pH Test gives 120 seconds. eLitmus writes that it moved away from a test challenged by time towards one challenged by content: it prints every formula you need in the paper and grades its questions by difficulty, so depth matters there more than speed.
• GATE’s 166 seconds is an average over the 1-mark and 2-mark questions of the whole paper, General Aptitude included.
• A slow habit costs whole questions. The last three lines show it: practising Logical Ability at 90 seconds a question means that 2 of its 12 questions are never read, and at 120 seconds, 5 of them.

A wrong answer costs nothing, a fixed share or an unpublished amount

Here the tests differ most, and the difference decides when you should guess. Under negative markingnegative marking: A scoring rule under which a wrong answer loses marks, often a fixed share of what a right answer earns. A question left blank usually scores zero. More in “Negative marking: when guessing pays” a wrong answer takes marks away, and each test sets its own price:

eLitmus says only that its penalty takes into account how many questions you attempt and how accurate those attempts are, and that a near miss counts differently from an unrelated wrong option. Because the formula is not public, nobody can work out a break-even point for guessing there; the safe reading is that accuracy counts for more than the number of questions attempted. CAT 2026’s press release does not cover marking, so read the rules on the CAT website before that test. The next lesson works out exactly when a guess pays under each of the published rules.

Marks, percentiles and best scores reward different attempts

Three kinds of score turn up, and each changes what a good attempt looks like:

• Marks added up. GATE and SSC CGL add up marks, and where a paper runs in several sessions or shifts, both convert the marks with a published normalisationnormalisation: Converting the marks of different sessions of one test onto a common scale, so that candidates who sat a harder session are not put at a disadvantage., which GATE’s brochure says allows for differences in difficulty between sessions.
• Marks weighted by difficulty and compared with other candidates. eLitmus says your pH Test score reflects how many questions you got right, how difficult they were and how other candidates did on the same questions, and it reports both an absolute score and a percentilepercentile: A score's place among the other candidates' scores. A score at the 90th percentile is higher than about 90 % of the scores it is compared with; it does not count right answers.. A percentile places you among the other candidates; it does not count your right answers, so a hard paper and an easy one can give the same percentile for very different marks.
• A score that shows ability, retaken at will. TCS iON sets no cut-off and no pass or fail for the NQT: each section gets a score that indicates your current ability. The number of attempts is not limited, the scorecard shows only your best score, and a score can be used for two years from the day its result is published.

AMCAT’s About page says it uses adaptive testing, without giving details. In an adaptive testadaptive test: A test that chooses each next question to suit the ability it has estimated from your answers so far: usually a harder question after a right answer and an easier one after a wrong one. the questions are chosen as you go. The official site of the ASVAB, a test that the armed forces of the United States give their applicants, explains the rule its adaptive version follows. Step through an invented candidate’s first five answers under that rule:

After each answer, the next question is picked to suit the ability estimated so far, usually an easier one after a wrong answer and a harder one after a right answer. Two candidates, or two attempts, need not see the same questions, so aim for steady accuracy on every question rather than speed on the first few.

Personality and coding parts are prepared differently

Some tests attach parts that are not aptitude at all, and each kind is prepared in its own way:

AMPI is based on the five-factor modelfive-factor model: A description of personality by five broad traits: extraversion, agreeableness, conscientiousness, neuroticism and openness to experience. of personality, which describes personality by five broad traits: extraversion, agreeableness, conscientiousness, neuroticism and openness to experience. Like any psychometric testpsychometric test: A questionnaire that describes traits such as personality, behaviour at work or work values, instead of testing knowledge with right and wrong answers., it describes you rather than marking you right or wrong, so read each statement and answer at the pace the module sets, about 13 seconds a statement. Coding and subject modules are the opposite case: they have right answers, and a syllabus to learn them from.

One original item for each NQT area, checked before you mark it

These are original items, one for each area of the TCS iON NQT Cognitive part, each labelled with the skill it practises. Try each before you read its answer.

1. Numerical, percentages. A laptop sells for ₹46,800 after a 10 % discount on its marked price. What is its marked price? (A) ₹51,480 (B) ₹51,800 (C) ₹52,000 (D) ₹52,800
2. Reasoning, number patterns. Which number comes next: 3, 4, 6, 10, 18, ? (A) 26 (B) 32 (C) 34 (D) 36
3. Verbal, usage. Choose the word that fits: “All candidates must comply ___ the dress code of the test centre.” (A) to (B) with (C) by (D) for

The answers: (C), (C) and (B). The program checks the first two and the trap answer of each:

// Checks the numerical and reasoning sample items of the lesson, and the trap answer of each.

// Numerical: a laptop sells for Rs 46,800 after a 10 % discount on its marked price. What is the marked price?
const sale = 46800;
const discount = 10; // per cent of the marked price
const marked = (sale * 100) / (100 - discount); // the sale price is 90 % of the marked price
const trap = sale + (sale * discount) / 100; // adds 10 % of the sale price instead
console.log(`marked price: ${marked}; the trap answer adds 10 % of the sale price: ${trap}`);
console.log(`check: ${marked} less 10 % is ${marked - (marked * discount) / 100}`);

// Reasoning: 3, 4, 6, 10, 18, ?
const series = [3, 4, 6, 10, 18];
const gaps = series.slice(1).map((x, i) => x - series[i]);
const doubling = gaps.every((g, i) => i === 0 || g === 2 * gaps[i - 1]);
const last = series[series.length - 1];
const next = last + 2 * gaps[gaps.length - 1];
console.log(`gaps: ${gaps.join(', ')}; each gap doubles: ${doubling}; next term: ${next}`);
// A second description of the same rule is a quick check: "double the term, then subtract 2".
const second = series.slice(1).every((x, i) => x === 2 * series[i] - 2);
console.log(`"double, then subtract 2" fits every step: ${second}; it also gives ${2 * last - 2}`);
// The two usual wrong answers: repeating the last gap, and doubling the last term.
console.log(`trap answers: ${last + gaps[gaps.length - 1]} (the last gap again), ${2 * last} (doubling)`);

marked price: 52000; the trap answer adds 10 % of the sale price: 51480
check: 52000 less 10 % is 46800
gaps: 1, 2, 4, 8; each gap doubles: true; next term: 34
"double, then subtract 2" fits every step: true; it also gives 34
trap answers: 26 (the last gap again), 36 (doubling)

• Item 1. The sale price is 90 % of the marked price, so the marked price is 46,800 × 100 ÷ 90 = ₹52,000. The trap, (A), adds 10 % of the sale price, but the discount was 10 % of the marked price, a larger amount. Check by going forwards: ₹52,000 less 10 % is ₹46,800.
• Item 2. The gaps between terms are 1, 2, 4 and 8: each gap doubles, so the next gap is 16 and the next term 34. A second description of the same rule, “double the term, then subtract 2”, fits every step and gives 34 too, which is a quick way to confirm a pattern. Repeating the last gap gives (A), and doubling the last term gives (D).
• Item 3. Oxford Learner’s Dictionaries gives the pattern “comply (with something)”: you comply with a rule, an order or a standard, so (B) is the only one that fits.

Each answer above has a second check: going forwards from the answer, or a second description of the same rule. Such a check takes a few seconds and catches the trap answers each item was built around.

Practise: read the rules, then work the pace

The questions cover the shared core, paces, penalties, scores, attached parts and the sample items. A session starts from the one you are least sure of and has no clock.

Interview questions

What do placement aptitude tests have in common, and which of their rules differ? (warm-up)

Almost all of them test the same three areas: quantitative ability (arithmetic, number work and data interpretation), logical reasoning (patterns, arrangements and deduction) and verbal ability (grammar, vocabulary and reading). Four rules differ, and each changes preparation. Time can be one clock for the paper or a fixed clock per section, as in CAT’s 40 minutes a section. Wrong answers may cost nothing, as in TCS iON NQT, a fixed share of the marks, as in GATE MCQs, or follow an unpublished rule, as in eLitmus. Scores can be plain marks, marks weighted by difficulty and compared with other candidates, or a best score you can improve by retaking the test. Some tests attach personality, coding or domain modules. A good answer turns each rule into a practice decision, such as training at the test’s seconds per question.

Key takeaways

• Nearly every placement aptitude test covers quantitative ability, logical reasoning and verbal ability, so you prepare by topic and then adapt to each test’s rules.
• Four rules differ from test to test: how the time is given, what a wrong answer costs, how the score is made and what comes attached.
• Seconds per question is the pace to practise at: 60 for AMCAT’s English module, 75 for its Logical Ability module, 36 for SSC CGL Tier-I and 120 on average for the eLitmus pH Test.
• A wrong answer costs nothing in TCS iON NQT, a fixed share in GATE’s MCQs and in SSC CGL, and an unpublished amount in eLitmus, where accuracy counts for more than attempts.
• Adaptive and peer-compared scores reward steady accuracy from the first question, and a personality module has no right answers to drill.