Why two of these sections behave differently
A quantitative section tests whether you know a method and can execute it quickly. Abstract reasoning tests something else: whether you can identify an unfamiliar rule under time pressure. There is nothing to memorise, because a question you have seen before is no longer testing the skill it was designed to test.
What does improve is process. Candidates who do well are not more insightful; they run through a fixed checklist - what is changing in count, position, orientation, size, shading - fast enough that the rule surfaces before the clock does. Critical reasoning works the same way, with a different checklist: locate the conclusion, then ask what must be true for it to follow.
What the assessment covers, and the caveat
The assessment has historically combined quantitative ability, verbal ability, abstract reasoning and critical reasoning, with some tracks adding a common-applications component covering basic computing and productivity tools. Accenture varies this by role and hiring cycle, so treat any pattern you find online - including this description - as indicative rather than current.
Confirm the sections and timing on the official Accenture careers portal for your application before planning around them. One practical note: the common-applications questions, where they appear, are pure recall and among the easiest marks available, so they are worth an hour if your track includes them.
The abstract reasoning checklist
Real abstract questions are visual - sequences of shapes where you predict the next figure. The text-based pattern questions further down train the same habit of testing rules systematically, but for the visual ones you need a checklist you can run in seconds. Work down it in this order and stop at the first thing that changes consistently.
| Check | What to look for | Typical rule |
|---|---|---|
| Count | number of shapes, sides, dots or lines | increases by one each step |
| Rotation | orientation of the whole figure or one element | 90° clockwise per step |
| Reflection | mirror image across a vertical or horizontal axis | alternates each step |
| Shading | filled, empty, half-filled, hatched | cycles over three steps |
| Position | an element moving around the figure | moves one corner clockwise |
| Size | one element growing or shrinking | alternates large and small |
| Addition/removal | an element appearing or disappearing | added every second step |
The five mistakes that cost the most marks
- ✓Memorising abstract question types. The section exists to test rule-finding on unfamiliar patterns, so recall of past patterns is exactly what it is designed to defeat.
- ✓Staying too long on one abstract question. If the rule has not appeared in about a minute it usually will not, and two easier questions elsewhere are worth more.
- ✓Treating critical reasoning as comprehension. Understanding the passage is necessary but not sufficient - the question is about the logic connecting its claims.
- ✓Skipping the common-applications questions where the track includes them. They are recall, they are quick, and they are the cheapest marks in the paper.
- ✓Practising untimed. The constraint is speed under pressure, and untimed practice measures a skill the assessment does not test.
Practice set 1: quantitative ability
1. Three-eighths of a number is 132. Find the number.
352. From (3/8)n = 132, multiply both sides by 8/3: n = 132 × 8/3 = 352. Checking backwards takes a second and is worth it - 352/8 = 44, and 44 × 3 = 132. Reversing a fraction of a number means multiplying by the reciprocal, and the most common slip is multiplying by 3/8 again instead.
2. The price of an item falls from 250 to 200. What is the percentage decrease?
20%. The decrease is 50, and percentage change is always measured against the original value: 50/250 = 20%. Dividing by 200 gives 25%, which answers a different question - by what percentage the new price must rise to return to 250. Both figures are correct for their own question, which is why identifying the base matters more than the arithmetic.
3. Find the simple interest on 6,400 at 12.5% per annum for 4 years.
3,200. Using PRT/100: 6400 × 12.5 × 4/100. Recognising 12.5% as 1/8 makes it mental - one-eighth of 6400 is 800 per year, and 800 × 4 = 3,200. The interest happens to be exactly half the principal here, which is the check that 12.5% × 4 years = 50% in total.
4. A car covers 240 km at 60 km/h and returns at 40 km/h. What is the average speed for the whole journey?
48 km/h. Average speed is total distance over total time, never the average of the speeds. The trip out takes 4 hours and the return 6, so 480 km in 10 hours gives 48 km/h. For equal distances the shortcut is the harmonic mean, 2 × 60 × 40/(60 + 40) = 48. The naive answer of 50 is too high because more time is spent at the slower speed.
5. Pipe A fills a tank in 12 minutes and pipe B empties it in 18 minutes. If both are open, how long does the tank take to fill?
36 minutes. Work in rates and let emptying be negative: 1/12 − 1/18 = 3/36 − 2/36 = 1/36 of the tank per minute, so filling takes 36 minutes. The sign convention is the whole of the method - if the outlet were faster than the inlet, the net rate would be negative and the tank would never fill, which is a legitimate answer to some versions of this question.
6. If 5 machines make 5 widgets in 5 minutes, how long do 100 machines take to make 100 widgets?
5 minutes. Divide down to one machine first: 5 machines making 5 widgets in 5 minutes means each machine makes one widget in 5 minutes. A hundred machines working in parallel therefore make a hundred widgets in the same 5 minutes. The instinctive answer of 100 minutes comes from scaling the total without noticing the machines scaled too.
7. Two dice are thrown. What is the probability that the sum is 8?
5/36. There are 36 equally likely outcomes, and five give a sum of 8: (2,6), (3,5), (4,4), (5,3) and (6,2). Note that (2,6) and (6,2) are distinct outcomes while (4,4) occurs once - treating the dice as indistinguishable is the error that produces wrong answers throughout two-dice probability.
8. A shopkeeper sells two items at 990 each, gaining 10% on one and losing 10% on the other. What is the net result?
A loss of 20, which is 1% of the total cost. The cost prices are 990/1.1 = 900 and 990/0.9 = 1,100, totalling 2,000 against a total selling price of 1,980. Equal percentage gain and loss at the same selling price always produce a net loss, because the loss applies to the larger cost price - the general result is a loss of x²/100 percent, here 1%.
9. In an examination, 35% failed in Mathematics, 25% failed in English and 10% failed in both. What percentage passed in both?
50%. The percentage failing at least one subject is 35 + 25 − 10 = 50, subtracting the overlap because those students were counted in both groups. The remaining 50% passed both. Forgetting to subtract the intersection gives 60% failing and 40% passing, which is the standard error in every set question of this shape.
10. What is the largest number that divides 60, 84 and 108 exactly?
12. This is the HCF. Factorising: 60 = 2² × 3 × 5, 84 = 2² × 3 × 7, and 108 = 2² × 3³. The common factors are 2² and 3, giving 12. Take the lowest power of each shared prime for the HCF and the highest power of every prime for the LCM - one rule, two directions, and mixing them up is the usual mistake.
Practice set 2: abstract and pattern reasoning
These are text versions of the rule-finding the abstract section tests visually. Work each one by proposing a rule and verifying it against every given term before answering - a rule that fits the first two terms and fails on the third is the most common way these are got wrong.
11. Find the next term: 2, 5, 10, 17, 26, ?
37. The differences are 3, 5, 7, 9 - consecutive odd numbers - so the next difference is 11, giving 37. The underlying rule is n² + 1 for n = 1, 2, 3 and so on, which is worth spotting because differences forming an arithmetic sequence always indicate a quadratic rule.
12. Find the next term: 4, 9, 25, 49, 121, ?
169. These are the squares of the primes: 2², 3², 5², 7², 11². The next prime is 13, so the answer is 169. The check that confirms the rule rather than a coincidence is the gap - 9 to 25 skips 16, which rules out consecutive squares and forces you to look at what the roots have in common.
13. If AZ, BY and CX follow a pattern, what comes next?
DW. The first letters advance A, B, C, D while the second letters run backwards from the end of the alphabet: Z, Y, X, W. Paired-letter series almost always work this way, with the two positions following independent rules, so separate them before looking for a pattern in either.
14. Find the odd one out: 121, 144, 169, 200, 225.
200. The others are 11², 12², 13² and 15². 200 is not a perfect square - 14² is 196, so the sequence also skips one, which is deliberate misdirection. Test the obvious classification first and do not assume a list is consecutive just because it is increasing.
15. Find the next term: 7, 26, 63, 124, ?
215. Each term is one less than a cube: 2³ − 1, 3³ − 1, 4³ − 1, 5³ − 1. The next is 6³ − 1 = 215. When terms sit just above or below round-looking values, test cube ± 1 and square ± 1 early - a large share of series questions are built on exactly that offset.
16. Complete the analogy: 8 : 512 :: 5 : ?
125. The relationship is cubing, since 8³ = 512, so 5³ = 125. Analogy questions require identifying the operation from the first pair and applying it unchanged - the trap here is reading 512 as a multiple of 8 and looking for a multiplier, which fits the first pair but has no consistent meaning.
17. In a code, RED is 27 and BLUE is 40. What is GREEN?
49. The code sums the alphabetical positions: R(18) + E(5) + D(4) = 27, and B(2) + L(12) + U(21) + E(5) = 40. For GREEN, 7 + 18 + 5 + 5 + 14 = 49. Sum-of-positions is one of three codes that account for most of these questions, alongside a fixed letter shift and reversal - test all three before inventing a fourth.
18. In the rows (4, 6, 24), (5, 3, 15) and (7, 8, ?), find the missing number.
56. The third entry is the product of the first two: 4 × 6 = 24 and 5 × 3 = 15, so 7 × 8 = 56. Matrix questions require the rule to hold in every complete row, so verify it against all of them before applying it - a rule that works on one row is a guess.
19. Find the next term: 2, 7, 4, 14, 6, 21, 8, ?
28. Two series alternate. The odd positions run 2, 4, 6, 8, increasing by 2, and the even positions run 7, 14, 21, increasing by 7 - so the next even-position term is 28. When a sequence rises and falls irregularly, splitting it into alternate terms is the first thing to try.
20. Find the next term: 1, 2, 6, 24, 120, ?
720. Each term multiplies the previous one by an increasing integer: ×2, ×3, ×4, ×5, so next is ×6 giving 720. These are the factorials, 1!, 2!, 3!, 4!, 5!, 6!. Checking the ratio of consecutive terms is what reveals this instantly, and it is the first check to run whenever terms grow faster than any polynomial would explain.
Practice set 3: critical reasoning
Each of these gives a short argument. Before reading the answer, identify the conclusion in your own words - most errors in this section come from attacking a claim that was never the point being argued.
21. "Our new packaging launched in March. Sales rose 15% in April. The packaging is clearly working." What is the conclusion, and what is the evidence?
The conclusion is that the packaging is working; the evidence is the 15% sales rise following its launch. Separating these is the first move in every critical reasoning question, and the signal words help - clearly, therefore, so and thus introduce conclusions, while because, since and after introduce evidence. The argument's weakness only becomes visible once you can see that it rests entirely on the sequence of two events.
22. In the argument above, what is being assumed without being stated?
That nothing else could account for the April rise. The argument treats the packaging as the only plausible cause, which requires an unstated premise ruling out seasonality, a price change, a marketing campaign or a competitor's problems. An assumption is something the argument needs to be true but never says, and the reliable test is negation: if a competing explanation does exist, the conclusion no longer follows.
23. Which of these would most weaken the argument: (a) the packaging cost more than expected, (b) sales rose 15% in April of each of the last five years, (c) some customers said they preferred the old packaging?
(b). If April always brings a 15% rise, the observed increase is exactly what would have happened anyway, and the evidence for the packaging's effect disappears. Option (a) concerns cost rather than effectiveness, and (c) is a minority preference that does not contradict a net rise. Weakening means undermining the link between evidence and conclusion, not merely saying something unfavourable.
24. Which would most strengthen the argument: (a) the design team is experienced, (b) sales fell 3% in a comparable region that kept the old packaging, (c) the packaging uses recycled materials?
(b). A comparable region without the change acts as a control, so a divergence between the two isolates the packaging as the cause. Options (a) and (c) are favourable facts about the packaging that say nothing about its effect on sales. Strengthening works the same way as weakening in reverse - it must bear on whether the evidence actually supports the conclusion.
25. "Students who attend the optional revision class score higher on average than those who do not. The class should be made compulsory." What is wrong with this reasoning?
It assumes the class causes the higher scores, when the students who choose to attend an optional class are likely to be more motivated to begin with. The score gap may reflect who selects into the class rather than what the class does, so making it compulsory would bring in exactly the students the correlation says nothing about. Whenever participation is voluntary, self-selection is the first alternative explanation to consider.
26. "We surveyed our customers and 92% rated our service highly, so most people in the market are satisfied with us." What is the flaw?
The sample cannot support the conclusion. Surveying only current customers excludes everyone who tried the service and left - the group most likely to be dissatisfied - so the 92% describes the people who stayed, not the market. The conclusion generalises to a population the sample was never drawn from, which is the defining feature of a sampling flaw.
Practice set 4: verbal ability
27. Choose the word closest in meaning to "ambiguous".
Unclear, in the particular sense of being open to more than one interpretation. The distinction matters: something vague lacks detail, while something ambiguous may be perfectly precise yet support two readings. "The chicken is ready to eat" is ambiguous rather than vague, and that difference is what the question is testing.
28. Spot the error: "The company have announced its results."
"Have" should be "has". The sentence already treats company as singular by using its, so the verb must agree - mixing a plural verb with a singular pronoun is the inconsistency. British usage does allow a plural verb for collective nouns when emphasising the members, but then the pronoun must follow suit: "the company have announced their results".
29. Give one word for "a word that reads the same forwards and backwards".
Palindrome. From the Greek palin, meaning again or back, and dromos, a running - so literally running back. Level, radar and rotator are examples. As with polyglot and its relatives, recognising the Greek and Latin fragments is what makes one-word-substitution questions fast rather than a matter of luck.
30. Choose the correct form: "She is one of those students who ___ always punctual."
Are. The relative clause describes those students, a plural noun, so the verb is plural - the sentence means she belongs to a group of students who are always punctual. Reading it as agreeing with she and choosing is is the standard error. The test is to find the noun the who actually refers to, which is the one immediately before it.