Frame the problem and protection gap Review questions Frame the problem and protection gap Registry coverage does not imply effective protection for everyone meeting a programme rule. Begin with denominator, period and counting unit: timely payment among enrollees differs from timely receipt among all eligible households. State uncertainty around estimated eligibility and missing cases, then locate the service-stage bottleneck. When a sponsor demands proof of reduced hardship without baseline or comparison, diagnose decisions and payments now and reserve impact claims for suitable design. Registries describe people inside them; a sound survey may reach non-applicants; interviews suggest mechanisms but cannot size them alone. Do not average rates with unlike definitions. Align them, name rival explanations and specify a prediction that separates each. In a case where files stall at document upload, a small assisted-access test may be warranted while the number of non-starters remains unknown.
Review questions Test your understanding with these questions. Marking and saved results are available with a free account.
Question 1 An agency wants proof that a benefit reduced hardship, but has three months of payment records and no hardship baseline. Which advisory question can be addressed now? How much hardship did the programme causally remove? Where does the transition from eligibility decision to timely payment fail? Would hardship end if benefit value doubled? Question 2 A cumulative registry says 80%, a quarterly survey 58%, and interviews come from one service centre. What first check makes comparison more defensible? Average them to 69% because both are numerical. Use the registry because it has more transactions than the survey. Align periods, enrolment definitions and eligible denominators, then examine sampling. Question 3 Using the lesson data, the agency wants the share of eligible households that received the benefit on time. Which calculation answers that question? 3,900/6,000=65% 3,900/4,800=81.25% 4,800/6,000=80% Question 4 There are 3,900 timely recipients among 4,800 enrollees. An official calls the 900-person difference rejected claims. What is the best advisory response when 300 were paid late? Classify all 900 as rejected because missing an on-time payment means rejection. Separate the 300 late payments from 600 cases with unknown status and request statuses and dates before interpreting the remainder. Remove the 300 late payments from the denominator to improve the on-time payment rate. Question 5 With a fixed budget, the choice is a categorical child benefit or income testing with stale data. What should be compared first? Only the number of files in the income-test system. Reach, value, exclusion risk and verification cost under each option. Only amount per person because administration cannot affect the choice. Question 6 Of 400 households meeting a programme rule, 280 receive support. What is exclusion under that rule? 120/400=30% 120/280≈42.9% 280/400=70% Question 7 Digital intake completes 75% and an assisted desk 85%, but desk decisions are slower. What is a suitable first follow-up? Declare the desk superior solely from higher completion. Close digital intake because it is lower on one metric. Trace valid completion, handoff and decision time by channel and case mix. Question 8 A 200-unit benefit arrives in half the months against a 500-unit monthly purpose-specific basket. What is a bounded adequacy description if prices stay fixed? It covers 40% of the basket every month of the year. It covers 40% in paid months and averages 20% across the year, with interruptions shown. The ratio proves the household is legally ineligible. Question 9 A committee weighted reach 40% and equity 25%, then new evidence reduced one option's equity score from three to two. What should follow? Recalculate ranking and document why the score changed. Keep the old total because weights were pre-agreed. Raise reach weight quietly to offset the lower equity score. Question 10 A campaign costs 150,000 and forecasts 400 additional timely recipients. What does 375 mean? The benefit amount each household received. Cost per forecast additional timely recipient under a defined horizon and cost scope, not a proved causal effect. The campaign's observed success percentage. Question 11 An agency approves 1,000 cases; its payment provider returns 920 successful settlements and 80 failures. Which count is confirmed paid? 1,000 because approval is enough evidence of receipt. 980 after assuming most failures will be fixed soon. 920, with a separate correction path for the eighty. Question 12 A control holds 100 transactions and twenty are true duplicates. Which hold-precision measure should accompany delay harm? 20/100=20% true among held, with eighty valid cases delayed. 20/80=25% because valid cases form the control denominator. 100% because every held case needs review. Question 13 Document assistance raises completed applications, but verification capacity remains below inflow. Which theory-of-change link needs review? Only the link from transfer to benefit value. Only outreach to awareness. Completion to timely decision; downstream capacity is a threatened assumption. Question 14 Of 1,000 households due payment, 780 are confirmed on time and forty await settlement. What can be reported now? 78% confirmed, 4% unresolved, upper 82% if all forty prove timely. 82% confirmed because reconciliation usually follows. Only 78%, omitting the forty so the indicator stays simple. Question 15 Hardship falls 40% to 30% for recipients and 35% to 30% for a comparison group. What is the difference in changes, in percentage points? −10 because that is the recipient change alone. −5, with causal attribution conditional on comparability. Zero because final rates are equal. Question 16 Overall reach gains ten points, but a group loses fifteen; the pilot rule bars a drop above five. What first decision follows? Scale because aggregate improvement met its trigger. End the underlying benefit for everyone immediately. Pause affected-channel expansion, check cases and cause, and preserve basic service. Question 17 Payment repair forecasts 600 additional recipients, but this was not measured. How should advice state it? Repair delivered to 600 extra households. It may reach 600 under a stated assumption, with confirmed count measured after a bounded pilot. Omit the number entirely because forecasts cannot inform decisions. Question 18 An agency has 100,000 recipients and raises value by 50 monthly over twelve unchanged cycles. What is extra transfer commitment before administration? 60 million units. Five million units for the whole year. Six hundred million units because there are twelve cycles. Question 19 A household challenges stale income in its record and service staff identify an urgent protection concern. How should roles be split? The policy adviser decides eligibility and protection together to avoid handoffs. Defer protection referral until the policy committee gathers complaint samples. Service routes the claim to eligibility authority and urgent concern to the specialist under approved channels; adviser receives aggregates only. Question 20 Of an estimated 10,000 eligible households, 8,000 enrol and 6,000 receive on time. What does 60% represent? Payment reliability among enrollees. Confirmed timely receipt among estimated eligible households, conditional on the estimate and period. The share of households denied the benefit. Create a free account to mark your answers and save your progress.