Problem diagnosis and decision boundaries Review questions Problem diagnosis and decision boundaries Regulatory analysis defines the problem before selecting an instrument. Separate an applicant's whole journey from processing after a file is complete; specify the population, denominator and clock start. Test rival explanations with stage data rather than one average. A baseline is not a frozen snapshot: include approved changes in the no-action trajectory and expose uncertainty instead of crediting their effect to a proposal. Map the obligation's rank and the body authorized to interpret or change it; a useful amendment does not create legal power. When records and interviews diverge, inspect coverage, measurement and selection. End with a brief stating testable facts, protection and burden criteria, preliminary options and a fact that could reverse the advice. Trial a limited step within confirmed authority while referring legal questions to the competent body.
Review questions Test your understanding with these questions. Marking and saved results are available with a free account.
Question 1 Training case: average permit time rises from eight to twelve days, but the new system starts the clock at first portal entry while the old system started after the application was complete. What should the analyst do first? Treat the four-day increase as proof that the rule failed Exclude incomplete applications only from the new year Recalculate both periods using one clock start and one completeness definition Compare staffing alone because the clock definition is merely technical Question 2 In a training case, 40 complete applications take six days each; 20 returned applications take eight days before completeness and six after. Which reading supports causal inquiry without distortion? The overall average is ten days because (6+14)÷2 averages the two groups The returned group takes 14 days and the overall average is about 8.67; examine the return stage separately The overall average is six days because the decision stage is shared The returned group takes eight days because decision time is unrelated to returns Question 3 A previously approved booking service starts next month, after which a regulatory amendment is proposed to cut waiting. How should the amendment's effect be compared? Include the expected but uncertain booking effect in the no-amendment path and compare the amendment with it Freeze current waiting time for a year so every later change is credited to the amendment Exclude booking from every option because it was approved earlier Count the combined effect of booking and amendment as the amendment's effect alone Question 4 A fictional text requires an annual safety check; department guidance specifies paper. A manager wants digital certificates but the team lacks delegation evidence. What is responsible advice? Abolish the annual check because digital access reduces burden Announce digital acceptance immediately and seek legal advice later Keep the check duty, obtain the text and delegation to test authority over proof format, and improve current guidance Reject all digital study because present guidance uses paper Question 5 Delay complaints fall from 80 to 50, but the complaint link becomes hidden. In interviews selected from returned files, 12 of 15 complain. What should the analyst conclude? Fewer complaints prove that the change succeeded Reporting opportunity and interview selection changed; repair measurement and inspect a wider sample before judging Twelve of fifteen proves that all applications failed Divide 50 by 15 to measure overall satisfaction Question 6 Twenty of 80 applications are returned for an unclear image. The team does not know if a text mandates it, and it may support identity checking. Which initial memo is strongest? Remove the image because returning 25% is unacceptable regardless of its purpose Hire decision staff because the return proportion measures processing capacity Say four days of post-completeness processing proves applicants face no problem Trial clearer image specifications, track returns and identity errors, and check the duty's basis before proposing removal Question 7 Half of returned files lack an understandable information item, while a protective check remains mandatory. Which option set permits a real comparison? Only a strict rule and an even stricter rule Baseline, clearer guidance, a binding amendment after authority checks, and a mix retaining screening Only abolish screening or abolish permits Treat a portal in three colours as three options Question 8 One option is reversible guidance; another is a paid certificate for everyone with unproven safety benefit. Where should deeper appraisal focus? Guidance colour because it is faster to test Equally on every possible cell regardless of impact Certificate cost and distribution, detection value and authority to impose it Fees alone, scoring safety at zero because data are missing Question 9 Among 300 annual applications, assumed returns fall from 30% to 20% after guidance whose first year costs 3,600 units. What is direct cost per return avoided if the forecast holds? 120 units because 30 returns are avoided 12 units because the denominator is all 300 applications 40 units because the denominator is 90 baseline returns 360 units because the rate falls by ten points Question 10 A form costs 4,000 units and prevents 20 to 50 returns; verification costs 6,000 and is estimated to prevent 40. What does sensitivity of cost per avoided return alone show? The form always wins because its total cost is lower Verification always wins because its prevented count is certain Both options are equal across the entire form range The form loses at 20 and wins at 50; preference switches near 27 prevented returns Question 11 A mandatory certificate receives the highest potential protection score, but authority to require it is unverified. How should the result be reported? Raise protection's weight until it overrides the authority gap Give it zero on every criterion to conceal discussion Make authority a gate and describe possible effects with a verification path rather than immediate implementation Assume agreement on its importance confers power to impose it Question 12 Sixty responses come from a large association and two from protection beneficiaries. Their directions are unstated. What is the first participation gap? The association supports the decision because it responded more Protection beneficiaries are sparsely heard; a channel is needed to understand the option's effect on them Multiply two responses by 30 to equalize representation Discard association replies because it has an interest Question 13 Which consultation question helps compare a proposed certificate with existing screening? Don't you agree that the modern certificate is necessary? Do you like the agency's service? What risk would the certificate detect beyond current screening, and what supports its effect and cost? How many people can submit the same sentence favouring the certificate? Question 14 Ninety identical messages oppose a fee and nine invoices show 15 units, while the official fee is ten. How should the gap be read? Ninety independent cost measurements of 15 The official fee is certainly false because invoices show 15 Average the two figures into a 12.5-unit fee without checking The nine invoices show a five-unit gap whose components need checking; identical messages are a campaign, not a cost sample Question 15 A worker describes procedural burden and asks not to be named, while a manager wants the whole file public. What is the better step? Separate using the information from publishing it and check disclosure rules with the competent officer before a safe summary Publish the file because consultation is always transparent Promise absolute confidentiality regardless of rules Discard the account because it is sensitive Question 16 Consultation reveals five documented false rejections of valid uploads. Which reply shows how participation changed the option? Thank you; the comment was considered We will repair the size limit or provide help and measure false rejection and support cost, without generalizing a rate from five cases We announce that every application will succeed after publication We remove the safety-proof duty for failed uploads Question 17 A portal accepts one certificate format although fictional text says 'evidence of a check' without naming a form. What first implementation test fits? Reject every alternative automatically Accept any file named certificate without verification Test normal, ambiguous-alternative and missing-proof cases, referring format interpretation to the competent body Remove the check duty from the screen Question 18 A system launches next week, but no official interpretation covers files completed before the update. What should happen? Apply the new rule retrospectively for uniformity Continue non-impactful preparation and hold affected automated decisions pending competent transition guidance Exempt every old file from all duties Make the system date the legal effective date Question 19 After all mandatory visits, two slots remain. Fictional risk scores for sites X=12, Y=10 and Z=6. What is the provisional plan? X and Y, checking data quality and sampling lower scores where authority allows Z alone because its score is lower X and Y, with penalties imposed before inspection Cancel any mandatory visit for Z Question 20 One body sends 14 files daily and the next processes 11; the queue starts at zero. What happens after five steady days? No queue because files were sent Eleven files accumulate Seventy files accumulate Fifteen files accumulate at the second body Question 21 Site A fixes a low-impact record error; site B has failed protective equipment and immediate danger. What distinction is sound? An earlier score mandates the same penalty for both Exempt B because its past record was good Verify A's correction and promptly refer B's observed danger through the authorized safety route Set B's final penalty in the analyst's paper Question 22 Thirty of 120 files return and screening detects two hazards. A new certificate's added benefit is unknown. Which advice is strongest? Mandate the certificate because two hazards prove its benefit Trial clearer forms within authority while retaining screening, measure returns and harm, and test the certificate before mandating Remove screening because returns are 25% Promise that guidance will remove every return Question 23 Twelve of 80 registered files return, while 15 distinct people fail before submission among 95 distinct cases overall. Which display is correct? Add the two rates and call them return rate Calculate 12÷95 because all 95 registered a file Treat pre-submission failure as proof that the file would return Registered returns are 15%; pre-submission failures about 15.79%, with different denominators Question 24 Returns fall from 30% to 20% as guidance, extra staff and pre-registration blocking launch together. What is the strongest direct conclusion? Guidance alone cut returns ten points Staff alone caused all improvement The registered-file indicator fell ten points after a package; each component's contribution is unproven Nothing changed because causality is unknown Question 25 After an interface update, three confirmed compliant files receive wrongful refusals. What review trigger fits? Urgently inspect the setting and correct files within authority while retaining proof duty Legally abolish proof duty by analyst decision Wait for a report in three years despite repeated error Treat publication as success despite refusals Question 26 In the integrated case, 32 of 160 registered files return and 16 attempts stop before submission, possibly including eight retries. What first reading is sound? Returns are 30% because 16 and 32 should be added Registered returns are 20%; deduplicate failed attempts before defining a broader journey denominator Pre-submission failure is zero because no file was registered All 16 certainly belong to different people Create a free account to mark your answers and save your progress.