Casework: study guide

Worked examples and pitfalls

The negation test finds assumptions; nothing else does: Argument: 'The council should fit bin sensors across the district. In the trial depot they cut collection trips by a fifth.' Candidate assumption A: 'The trial depot's waste pattern is broadly representative of the rest of the district.' Negate it - suppose the trial depot is nothing like the rest of the district - and the argument collapses, because a result that does not transfer supports nothing about the district. So A is a required assumption. Candidate assumption B: 'Bin sensors are the cheapest available technology for this purpose.' Negate it - suppose they are the most expensive - and the argument still stands as given, since it argued from a reduction in trips, not from cost. B is not required; it would strengthen a cost-based argument that was never made. Candidate C: 'Fewer collection trips reduce total operating cost.' Negate it and the recommendation loses its point, so C is required too, even though the argument never mentions cost. That last case is the one candidates miss, because the required assumption is what bridges the evidence to the recommendation, and bridging assumptions are by definition unstated. The failure mode across this whole item type is picking statements that support the conclusion instead of statements the argument cannot survive without.

Base rates: 90 percent accurate, 15 percent right: A production line runs an automated defect test. Defects occur in 1 percent of units. The test flags 90 percent of genuinely defective units, and wrongly flags 5 percent of good units. A unit has just been flagged - how likely is it to be defective? Work in whole units rather than probabilities. Take 10,000 units: 100 are defective and 9,900 are good. Of the 100 defective, the test flags 90. Of the 9,900 good, it wrongly flags 5 percent, which is 495. Total flags: 90 + 495 = 585. Of those, 90 are genuine, so the answer is 90 / 585 = 15.4 percent. Better than the 1 percent base rate, and nowhere near the 90 percent that intuition offers. The error has a name - confusing the probability of a flag given a defect with the probability of a defect given a flag - and it has a practical consequence: the rework queue has to be sized for 585 units a batch, not 100, and 495 of the 585 units sitting in it are perfectly good. Whenever a test, screening rule or model accuracy figure appears in an item, ask how many of the negatives there are, because with a rare condition the false positives from a large clean population swamp the true positives from a small affected one.

Three alternative explanations, and the comparison that settles it: Claim: 'Employees who attended the optional resilience workshop took 30 percent fewer sick days last year, so the workshop works.' Alternative one, selection: attendance was optional, so the people who signed up may have been the healthier and more engaged to begin with, and would have taken fewer sick days regardless. Alternative two, reverse direction: employees who were frequently unwell were the least able to attend a full-day workshop, so illness determined attendance rather than the other way round. Alternative three, a common third factor: the workshop ran on Thursday afternoons at head office, so attendance largely tracks being an office-based rather than shift-based worker, and shift workers take more sick leave for reasons that have nothing to do with resilience training. What would settle it: randomly assigning the offer to half the workforce and comparing the two groups. Failing that, the cheapest useful check already exists in the payroll data - compare attendees' and non-attendees' sick days in the year before the workshop. If attendees were already 30 percent lower, selection is doing the entire job. The habit worth building is to name the comparison group before you accept any before-and-after number.

Deduction: affirming the consequent versus modus tollens: Rule: 'If an invoice is over 10,000 euro, it requires two signatures.' Argument A: 'This invoice has two signatures, therefore it is over 10,000 euro.' Invalid - this is affirming the consequent. The rule makes two signatures necessary for large invoices; it says nothing that prevents a cautious manager from double-signing a 400 euro invoice, so a two-signature invoice of any size is consistent with the rule. Argument B: 'This invoice has only one signature, therefore it is not over 10,000 euro.' Valid - this is modus tollens, denying the consequent to deny the antecedent, and it holds assuming the rule was followed. Argument C: 'This invoice is not over 10,000 euro, therefore it does not have two signatures.' Invalid again, denying the antecedent. Two further points that deduction items test directly. First, validity is about form alone: 'All banks close on Sundays; Riverton Mutual is a bank; therefore Riverton Mutual closes on Sundays' is perfectly valid even if the first premise is false, and a valid argument with a false premise can deliver a false conclusion. Second, in these items you must accept the premises as given even when you know them to be untrue, because the question is whether the conclusion follows, not whether it is true.

Interpretation: what a survey figure does and does not license: Data: all 1,200 employees of a firm were surveyed. 62 percent said they would accept a four-day week at 90 percent pay. Of those who said yes, 71 percent were under 35. Conclusion one: 'A majority of this firm's employees would accept the trade.' This follows - 62 percent of a complete census of the firm is a majority, and the conclusion is properly limited to this firm. Conclusion two: 'A majority of the firm's under-35 employees would accept the trade.' This does not follow, and the arithmetic shows why. Sixty-two percent of 1,200 is 744 accepters, and 71 percent of 744 is about 528 accepters aged under 35. That is the composition of the accepters, not the acceptance rate within the under-35 group, and without knowing how many under-35s the firm employs the rate is undetermined: if there are 900, the rate is 528 / 900 = 59 percent; if there are 600, it is 88 percent; if there were 1,100 the rate would be below half. Conclusion three: 'The firm should move to a four-day week.' This does not follow either - a stated preference about pay says nothing about output, shift cover, customer hours or cost. Notice that conclusion two is the base-rate error from the screening example, wearing survey clothing.

Strong and weak arguments on the same question: Question: should the agency publish raw inspection scores for every site? Strong argument for: 'In the neighbouring authority, publication was followed by a measurable fall in repeat violations, which is the stated aim of the inspection regime.' It is directly about this decision, it addresses the regime's own objective, and it offers evidence - and if true it would change your view. Weak argument for: 'Yes, because transparency is always good.' Sweeping, unsupported, and it would apply identically to publishing anything at all, which is the tell. Strong argument against: 'Raw scores omit the severity weighting, so a site with one critical failure can rank above a site with six minor ones, and a member of the public cannot see the difference.' Specific, mechanistic, directly about the proposal, and it would change your view. Weak argument against: 'No, because businesses will object.' It may be true and it may matter politically, but it does not bear on whether publication achieves the regime's aim, and no evidence is offered. Two tests separate strong from weak: is the argument about the exact question asked, and would accepting it change the decision? Note that you can hold a firm personal view on publication and still be required to mark the argument on the other side as the strong one - that split is what the item type is measuring.

How to practise this skill

• Negate every candidate assumption out loud. If the argument survives the negation, it was never an assumption, however supportive it sounds. This single habit converts the assumption item type from guesswork into a mechanical check.
• Write the conclusion in your own words before reading any option. Half of the wrong answers on inference and evaluation items are responses to a conclusion the argument never reached.
• Ask 'percentage of what?' on every percentage in the stimulus and reconstruct the denominator in whole units. Base-rate and composition errors both dissolve the moment you write out counts instead of rates.
• Deliberately practise marking arguments you disagree with as strong and arguments you agree with as weak. Assessments in this construct are built to catch agreement masquerading as evaluation, and the effect is largest on politically loaded stimuli.
• For any causal claim, write two alternatives - selection and a third factor - and name the comparison group that would rule them out, before you decide whether the evidence supports the claim.
• Log wrong answers by task type rather than by topic: inference, assumption, deduction, interpretation, evaluation. Candidates are rarely weak across all five, and the profile tells you where the next hour of practice belongs.

Glossary

Conclusion

The claim an argument is trying to establish. It is not always last, is often signalled by therefore, so, or should, and locating it correctly determines every subsequent judgement about the argument.

Assumption

An unstated premise the argument requires in order to work. Identified by negation: negate it and a genuine assumption brings the argument down, while a merely helpful statement leaves it standing.

Validity and soundness

An argument is valid when the conclusion cannot be false while the premises are true, which is a property of form alone. It is sound when it is valid and the premises are actually true.

Affirming the consequent

The invalid pattern 'if P then Q; Q; therefore P'. It is the most common deductive distractor because it differs from the valid modus tollens form by only the position of a negation.

Base rate

How common something is in the population before any test or evidence is applied. Ignoring it makes accurate-sounding tests appear far more informative than they are when the condition is rare.

Confounder

A third factor associated with both the supposed cause and the outcome, capable of producing the entire observed relationship on its own. Ruled out by randomisation or by an explicit comparison group.

Selection effect

A difference between groups created by how people entered them rather than by the treatment under study. Optional programmes and voluntary surveys are where it appears most often.

Falsifiability

The property of a claim that some observation could show it to be wrong. A claim compatible with every possible result carries no information, which is why 'what would change your mind?' is a diagnostic question.

Study this next

• Critical thinking skill hub
• Practise the critical thinking bank
• Verbal reasoning lesson
• Government audit suite
• Investigator and detective reasoning suite
• Critical thinking lesson on Novus Learn

Where this material comes from

• Worked items written for Novus Learn. The arguments, survey figures, defect-test numbers and evaluation options above are original and invented for this lesson; no published or copyrighted test item is reproduced.
• The five task types named in the summary - inference, assumption, deduction, interpretation, evaluation of arguments - describe a structure used publicly across several critical-thinking assessments; no affiliation with any publisher is claimed or implied.
• Terminology follows standard, widely published usage in introductory logic and research methods - validity, soundness, affirming the consequent, base rate, confounder, selection effect.
• Novus Learn aptitude construct registry (catalog seed) for the construct scope and the suite mapping shown in the related links.
• Public educational framing only - not affiliated with any official exam board, publisher or employer.

Worked examples and pitfalls

Agreement: find the head noun, then decide: 'The set of quarterly returns filed by the three regional offices (was / were) incomplete.' The subject is 'the set', which is singular, so 'was' is correct; 'offices' is merely the nearest noun, sitting inside a prepositional phrase that cannot supply the subject. Strike out everything between the head noun and the verb and the choice becomes obvious - 'the set was incomplete'. Next: 'Neither the manager nor the two supervisors (was / were) informed.' With neither/nor the verb agrees with the nearer of the two subjects, and the nearer one here is 'the two supervisors', so 'were' is correct. Reverse the order - 'Neither the two supervisors nor the manager was informed' - and the correct verb flips to singular, which is a useful check that you are applying the rule rather than guessing. Finally, a pair that trips experienced writers: 'A number of applications are missing' is correct, because 'a number of' functions as a plural quantifier, while 'The number of applications is rising' is also correct, because there 'the number' is a single quantity. Same three words, opposite verbs.

Dangling modifiers and the wandering 'only': 'Having reviewed the tender documents, the deadline was extended by two weeks.' An opening participial phrase attaches to the subject of the main clause, and the subject here is 'the deadline', so the sentence says the deadline reviewed the documents. It reads acceptably aloud because a listener silently supplies the missing actor, which is precisely why this error survives proofreading. Two clean fixes: name the actor as the subject - 'Having reviewed the tender documents, the panel extended the deadline by two weeks' - or drop the participle entirely - 'The panel reviewed the tender documents and extended the deadline by two weeks.' Now the limiting word. 'We only sent the invoice to three suppliers' literally says that sending was the only thing we did to those three. 'We sent the invoice to only three suppliers' says we sent it to no more than three. 'We sent only the invoice to three suppliers' says we sent nothing else. Three placements, three meanings, one word - put 'only', 'just', 'even' and 'almost' immediately before what they limit, and in a marked item the four options usually differ by nothing else.

Parallelism in a list: 'The role involves drafting reports, client liaison, and to maintain the case register.' Three items in three different grammatical forms: a gerund phrase, a noun phrase, and an infinitive. The repair is to pick one form and hold it: 'The role involves drafting reports, liaising with clients, and maintaining the case register.' Parallelism is worth its own item type because it is the most visible marker of careless list writing, and because real documents fail on it constantly - a bulleted list where two bullets start with a verb and the third starts with a noun, or a set of headings mixing questions and statements. The same rule governs correlative pairs: 'not only ... but also' must join two grammatically matching elements, so 'The report not only covers spending but also staffing' is faulty, and 'The report covers not only spending but also staffing' is correct, because the correlatives now bracket two nouns rather than a verb phrase and a noun.

Comma splice, and why 'however' is not a rescue: 'The system was offline for six hours, no backup was available.' Two independent clauses joined by a comma is a comma splice, and there are four legitimate repairs. Use a full stop: 'The system was offline for six hours. No backup was available.' Use a semicolon, which signals the two facts belong together. Add a coordinating conjunction after the comma: '... six hours, and no backup was available.' Or subordinate one clause: 'Because no backup was available, the six-hour outage could not be mitigated.' The trap is the fifth option that people reach for instinctively: 'The system was offline for six hours, however no backup was available.' However is a conjunctive adverb, not a coordinating conjunction, so it cannot join two independent clauses with a comma - the sentence is still spliced. It needs a semicolon or full stop before 'however' and a comma after it. The same applies to therefore, moreover, nevertheless, and consequently, which is why they appear so often in sentence-correction distractors.

Plain language: 37 words down to 14: Before: 'It is the recommendation of this department that consideration be given to the deferral of the implementation of the new filing procedure until such time as adequate training provision has been made available to all affected staff.' That is thirty-seven words. After: 'We recommend delaying the new filing procedure until all affected staff have been trained.' Fourteen. Look at what the rewrite actually did: it turned three nominalisations back into verbs (recommendation to recommend, deferral to delaying, provision to trained), it named the actor instead of hiding behind 'this department', and it cut 'consideration be given to' and 'until such time as', which carry no content at all. Look just as carefully at what it kept: the recommendation, the thing being delayed, and the condition that ends the delay. Plain-language items are marked on meaning preservation, not on brevity alone, and the standard wrong answer is the over-cut version - 'We recommend delaying the new procedure' - which reads beautifully and silently deletes the condition that is the substance of the recommendation. Trim the packaging, never the payload.

One incident, three readers, and the fact-inference line: Facts: the gate camera timestamps a van leaving at 11:42:00; the weighbridge camera, 300 m along the yard road, timestamps the same van at 11:42:15; the yard limit is 20 km/h. The arithmetic: 300 m in 15 s is 20 m/s, and 20 x 3.6 = 72 km/h. Now write it three ways. In the incident record: 'Vehicle VT-118 passed the gate camera at 11:42:00 and the weighbridge camera at 11:42:15. The cameras are 300 m apart, giving an average speed of 72 km/h over that section. The posted yard limit is 20 km/h.' Note what is absent: 'the driver was speeding', which is a conclusion, and 'was travelling at 72 km/h', which overstates the evidence - the figure is an average over the section, not an instantaneous speed. To the yard supervisor: lead with the decision needed and the deadline. 'Two camera timestamps put VT-118 at an average 72 km/h in a 20 km/h yard yesterday morning. I have attached the record. Can you confirm by Friday whether this goes to the safety review?' To the driver's employer: lead with the specific, verifiable reference - vehicle, date, time, the two camera points - and the single action requested. Same facts every time; what changes is which fact leads, and in a written-communication assessment the most apologetic option is rarely the best one, while any option that names blame in a record that has not been reviewed is usually the worst.

How to practise this skill

• Proofread in separate passes, one error class at a time - agreement, then modifiers, then punctuation, then numbers and names. A single pass looking for 'anything wrong' reliably finds fewer errors than three narrow passes taking the same total time.
• Read your draft aloud, or subvocalise it. Comma splices, dropped words and lost parallelism are audible even when they are invisible; dangling modifiers are the exception, which is why they need a deliberate check of what each opening phrase attaches to.
• On sentence-correction items, diagnose before you choose. Name the fault in the original in three words, then look for the option that fixes exactly that and introduces nothing new. Options that fix the fault while adding a second one are standard.
• Read every option to the end. Sentence-correction distractors commonly differ from the correct answer only in the last two or three words, and candidates who stop once the opening looks right lose marks they never see.
• For free-text tasks, spend the first minute on a three-line plan: purpose, the single action you want from the reader, and the deadline. Untimed free-text answers are usually well written and structurally shapeless, and structure is what the mark scheme can see.
• Keep a personal error list of the mistakes you actually make rather than a generic grammar checklist. Most writers repeat three or four faults, and proofreading against your own four is far faster than proofreading against forty.

Glossary

Subject-verb agreement

The requirement that the verb match the grammatical number of the subject's head noun, not of the nearest noun. Long prepositional phrases between the two are the standard way this is tested.

Dangling modifier

An introductory phrase whose implied actor is not the subject of the main clause, so the sentence attributes the action to the wrong thing. Fixed by naming the actor as the subject or by rewriting the phrase as a full clause.

Parallelism

Matching grammatical form across items in a list or across correlative pairs such as not only / but also. The commonest breach mixes gerunds, nouns and infinitives in a single series.

Comma splice

Two independent clauses joined only by a comma. Repaired with a full stop, a semicolon, a coordinating conjunction, or subordination - never by inserting however, therefore or moreover after the comma.

Nominalisation

A verb converted into a noun - implement becomes implementation, decide becomes decision. Nominalised prose hides who does what and typically doubles the word count for the same content.

Register

The level of formality and the vocabulary appropriate to a particular reader and purpose. A message can be grammatically perfect and still wrong in register, which is what free-text tasks are largely marking.

Active and passive voice

Active names the actor before the verb; passive can omit the actor entirely. Passive is legitimate when the actor is unknown or genuinely irrelevant, and evasive when it is used to avoid naming a responsible party.

Observation versus inference

An observation is what a camera, meter or witness recorded; an inference is the conclusion drawn from it. Records that mix the two are hard to defend later, because the conclusion cannot be separated from its basis.

Study this next

• Written communication skill hub
• Practise the written communication bank
• Reading comprehension lesson
• Police report writing suite
• Administrative and clerical entry suite
• Written communication lesson on Novus Learn

Where this material comes from

• Worked items written for Novus Learn. Every faulty sentence, rewrite, word count and incident figure above is original and invented for this lesson; no published or copyrighted test item is reproduced.
• Grammatical terminology follows standard, widely published usage - subject-verb agreement, dangling modifier, parallelism, comma splice, nominalisation.
• Novus Learn aptitude construct registry (catalog seed) for the construct scope and the suite mapping shown in the related links.
• Public educational framing only - not affiliated with any official exam board, publisher or employer.

Worked examples and pitfalls

Worked scenario: an unverified safety concern: A colleague reports a possible equipment fault while a deadline is approaching. First distinguish the known fact, the report, from the unverified cause. A strong response protects people and affected work, checks the concern through the right channel, tells the relevant lead, and records what was done. Ignoring the report underreacts; shutting down unrelated work or accusing someone before checking the facts overreacts.

Method: facts, duties, risks, response: Write four short notes before ranking options: what is known, who may be affected, which duty or boundary applies, and what safe next step is available now. Prefer an action that addresses the immediate issue and creates useful follow-through. Do not reward an option merely because it sounds decisive.

How to practise this skill

• Answer the question asked: best first action, worst action, or complete response are different tasks.
• Check whether an option acts within the person's authority and escalates only as far as the risk requires.
• When two options look reasonable, prefer the one that gathers missing facts and communicates ownership.

Glossary

Proportionality

Matching the urgency and scope of a response to the evidence, likely impact, and authority available.

Role boundary

The limit of what a person may decide or do without approval, specialist help, or escalation.

Follow-through

Confirming ownership, recording the decision, and checking that the issue was actually resolved.

Study this next

• Situational judgement test
• Workplace ethics
• Decision-making
• Situational judgment lesson on Novus Learn

Where this material comes from

• Novus Learn original situational-judgment suite scenarios and published construct mapping.
• Novus educational framework: facts, duties, risks, role boundaries, proportional action, and follow-through.

Worked examples and pitfalls

Worked scenario: a four-hour shutdown: Inspection takes 30 minutes before either repair can start. Two repairs take 90 and 120 minutes, but both need the same technician; testing takes 45 minutes after both finish. The minimum sequence is 30 + 90 + 120 + 45 = 285 minutes, longer than the shutdown. The correct response is to surface the conflict and change scope, staffing, or window, not to place overlapping bars on a chart and call the plan feasible.

Dependency and contingency check: Draw arrows between tasks, then mark scarce resources and approval gates. For the highest-risk dependency, define an early warning and a response: for example, if a permit is not approved by noon, defer nonessential work and notify the owner rather than discovering the conflict at start time.

How to practise this skill

• Check resource conflicts as well as task order; parallel work is impossible when both tasks need the same person or asset.
• Protect the completion milestone with visible review points, not an unexplained buffer at the end.
• When a constraint makes the plan impossible, state it early and propose the smallest viable change.

Glossary

Dependency

A rule that one task or milestone must precede or enable another.

Critical path

The dependent task sequence that determines the earliest possible finish under the stated plan.

Slack

The time a task may move without delaying a dependent milestone or final completion.

Study this next

• Scheduling
• Critical path method
• Resource allocation
• Planning and scheduling exercises lesson on Novus Learn

Where this material comes from

• Novus Learn original scheduling, project-coordination, dispatch, and operations suite scenarios.
• Novus educational framework: tasks, durations, dependencies, resources, constraints, milestones, and contingencies.

Worked examples and pitfalls

Two price structures and the distance where they cross: A delivery of 55 km. Courier A charges 8.00 base plus 0.45 per km. Courier B charges 20.00 flat for the first 40 km, then 0.90 per km beyond that. Which is cheaper? Courier A costs 8 plus 0.45 times 55, which is 8 plus 24.75, giving 32.75. Courier B costs 20 plus 0.90 times 15, which is 20 plus 13.50, giving 33.50. Courier A wins by 0.75. Now the question the item is really testing: is A always cheaper? At 45 km, A costs 8 plus 20.25 which is 28.25, while B costs 20 plus 4.50 which is 24.50. B wins comfortably. So the answer flips somewhere between, and finding where is one line of algebra. Above 40 km, B costs 20 plus 0.9 times distance minus 40, which simplifies to 0.9d minus 16. Set that equal to A's 8 plus 0.45d: 8 plus 0.45d equals 0.9d minus 16, so 24 equals 0.45d, so d is about 53.3 km. Below 53.3 km B is cheaper; above it A is. Check at the crossover: A is 8 plus 24 which is 32.00, and B is 48 minus 16 which is also 32.00. The general lesson is that whichever option has the lower per-unit rate always wins eventually, regardless of the base charges, and the base charges only decide where eventually starts. A single quoted distance never answers a which-is-cheaper question for a fleet.

Halving the search space beats walking the chain: Forty sensors sit on a single daisy-chained cable and exactly one connection is broken, cutting off everything downstream. How many tests do you need in the worst case? Walking the chain from one end and testing each sensor in turn takes up to 40 tests and 20 on average. Test the midpoint instead. If sensor 20 responds, the break is in the upper half and you have eliminated 20 candidates in one measurement; if it does not, the break is below and you have eliminated the other 20. Repeat on the surviving half: 40 becomes 20, then 10, then 5, then 3, then 2, then 1. Six tests, worst case, because each test halves what is left and two to the power six is 64, comfortably more than 40. The saving grows as the problem grows: 1,000 candidates need only ten tests. Two conditions have to hold for bisection to be valid, and stating them is part of the answer. The fault must be monotone, meaning everything on one side of the break behaves differently from everything on the other; and a single test at any point must tell you which side you are on. When there are two independent breaks, or when a test is only meaningful at the ends, bisection does not apply and a different strategy is needed. Candidates who reach for bisection reflexively on a problem that fails those conditions lose more than they save.

Working backwards from the number you were given: A team started a quarter with an unknown budget. In week one it spent half of it. In week two it spent 400 of what remained. In week three it spent a third of what remained after that. It finished with 1,200. How much did it start with? Attempting this forwards means carrying an unknown through three operations. Backwards it is arithmetic. After week three, 1,200 is what is left having spent a third, so 1,200 represents two thirds of the week-three opening balance, which was therefore 1,800. Before week two's spend of 400, the balance was 1,800 plus 400, which is 2,200. That 2,200 is what remained after spending half, so the starting budget was 4,400. Verify forwards, always: 4,400 less half is 2,200; less 400 is 1,800; less a third of 1,800, which is 600, leaves 1,200. Correct. The mechanical rule is to invert each operation and apply the inversions in reverse order. The inverse of spending a third is dividing by two thirds, not multiplying by three, and that specific slip is the most common wrong answer in this family. Working backwards is the right tool whenever the end state is known exactly and the operations are individually invertible, which covers most budget, mixture and journey problems phrased as how much did it start with.

When an estimate is enough, and when it is not: A maintenance window is three hours. Two technicians must service 1,850 units, each taking about 4.5 minutes, plus a shared 20-minute setup and 15-minute teardown. Feasible? Estimate first: 1,850 units at 4.5 minutes is 8,325 technician-minutes; split between two people that is about 4,163 minutes, which is roughly 69 hours. The window is 3 hours. The answer is no by a factor of more than twenty, and no refinement of the setup and teardown figures could change it. Precision here would be wasted effort. Now the same problem with 90 units. Ninety at 4.5 minutes is 405 technician-minutes, halved to 202.5 minutes, plus 35 minutes of setup and teardown, giving 237.5 minutes. That is 3 hours and 57 and a half minutes against a 3-hour window. Still no, but only just, and now every assumption matters: whether the technicians can genuinely work in parallel, whether setup is shared or duplicated, whether 4.5 minutes is a mean or a best case. The judgement being assessed is knowing which regime you are in. If your rough figure misses the target by an order of magnitude, stop and answer. If it lands within a factor of two, the estimate has not settled anything and you must do the exact arithmetic and name your assumptions.

The binding constraint decides, and it is rarely the obvious one: A van carries at most 900 kg and at most 6 cubic metres. Pallet type X weighs 150 kg, occupies 0.8 cubic metres and is worth 200. Pallet type Y weighs 90 kg, occupies 1.4 cubic metres and is worth 260. What mix maximises value? The instinctive heuristic is value per kilogram: X gives 200 over 150, about 1.33, while Y gives 260 over 90, about 2.89, so load Y. That is wrong, and the reason is that weight is not what runs out. Value per cubic metre tells the opposite story: X gives 250 and Y gives about 186. Enumerate the feasible whole-pallet mixes. Six X uses 900 kg and 4.8 cubic metres, worth 1,200. Five X and one Y uses 840 kg and 5.4 cubic metres, worth 1,260. Four X and two Y uses 780 kg and exactly 6.0 cubic metres, worth 1,320. Three X and three Y needs 6.6 cubic metres and does not fit. Two X and three Y uses 570 kg and 5.8 cubic metres, worth 1,180. Four Y alone uses 5.6 cubic metres and is worth 1,040. The best mix is four X and two Y at 1,320. Look at what binds it: volume is used to the last cubic metre while 120 kg of payload goes unused. Once volume is identified as the binding constraint, X's superior value per cubic metre explains the X-heavy answer, and the spare weight explains why two Y still get on board. The transferable move is to compute the usage of every constraint at your proposed answer and see which one is exhausted; a per-unit ratio computed against a non-binding resource is worse than no heuristic at all.

Cause, symptom, and the test that tells them apart: A pump trips out twice a shift. The obvious fix is to reset it, which works for about four hours. Ask why once and you find the thermal overload relay is tripping. Ask again and the motor is running hot. Again and the bearing is running hot. Again and the grease has dried out. Again and the lubrication interval was set for a duty cycle far lighter than the pump has actually been running since the line was rebalanced last year. Each level supports a different fix: resetting the trip costs nothing and lasts four hours; replacing the relay costs a part and lasts until the bearing seizes; regreasing lasts weeks; changing the lubrication schedule to match the real duty cycle is the one that stops the fault recurring. Stop asking why when the next answer stops being something you can act on, or when it leaves the boundary of the system you can change. There is one discipline that separates this from storytelling. A correlation is a lead, not a cause: the fact that the trips started after the line was rebalanced is suggestive, not proof. The confirming test is to intervene and observe: restore the original duty cycle, or apply the corrected greasing interval to this pump and not to the identical one on the parallel line, and see which one trips. If a proposed cause cannot be tested by changing it, treat it as a hypothesis and say so rather than closing the investigation.

How to practise this skill

• Write the question you are actually being asked at the top of your working, in your own words. Problem-solving items usually supply more information than the question needs, and the commonest failure is a correct calculation of something nobody asked for.
• Before any comparison of two options, ask where they cross. One quoted quantity never settles which is cheaper or faster, and the break-even is usually a single line of algebra away.
• State the worst-case number of steps before starting a search or a fault trace. If bisection applies, say so and say why; if the fault is not monotone or a test only works at the ends, say that instead and pick a different method.
• Estimate before you compute, then decide explicitly which regime you are in. Missing the target by an order of magnitude ends the question; landing within a factor of two means the estimate proved nothing and the exact numbers are now compulsory.
• Finish every allocation or capacity problem by listing how much of each constraint your answer consumes. The constraint you exhaust is the one that explains the answer, and the one you do not is where a per-unit heuristic will mislead you.
• Always verify a backwards solution by running it forwards. Attempts stay on this device, so the extra minute costs nothing and catches the inverted-fraction error that makes this family's most attractive wrong answer.

Glossary

Break-even point

The value at which two options cost or take the same. Below it one option wins and above it the other does, which is why a single quoted case never answers a comparison question.

Binding constraint

The limit that is fully consumed by the chosen solution. Identifying it explains the answer and reveals which per-unit ratios are relevant and which are noise.

Slack

The unused portion of a constraint at a proposed solution. Spare capacity in one resource is what lets a mix include items that look inefficient on the binding resource.

Bisection

Halving a search space with each test. It reduces a search of n candidates to about log base two of n tests, but only where the fault is monotone along the chain.

Working backwards

Solving from a known end state by inverting each operation in reverse order. The reliable method whenever the final value is given and every step is individually invertible.

Order-of-magnitude estimate

A rough calculation intended to settle feasibility rather than produce a figure. It is decisive when the answer is off by a factor of ten and useless when the margin is within a factor of two.

Root cause

The condition whose removal stops the fault recurring, as opposed to a symptom whose treatment suppresses it temporarily. A candidate cause that cannot be tested by changing it remains a hypothesis.

Monotone fault

A fault where everything on one side of a break behaves differently from everything on the other. The precondition that makes bisection valid, and the assumption most often left unstated.

Study this next

• Problem solving skill hub
• Problem solving practice bank
• Diagrammatic reasoning lesson
• Maintenance technician suite
• Supply chain planner suite
• Problem solving lesson on Novus Learn

Where this material comes from

• Every figure above was computed and checked for Novus Learn: the courier break-even at about 53.3 km, the six-test bisection bound for 40 candidates, the 4,400 starting budget verified forwards, and the four-X-two-Y loading enumerated across all feasible whole-pallet mixes.
• Terminology checked against the public Wikipedia articles 'Binary search algorithm', 'Break-even', 'Shadow price' and 'Root cause analysis'. Definitions only; every scenario above is original.
• Novus Learn aptitude construct registry (catalog seed) for the construct scope and related suite mapping.
• Public educational framing only: not affiliated with any official exam board, publisher or employer, and no copyrighted test item is reproduced.