01. Turn a sample target into a feasible fieldwork plan
Lesson
LEARNING OBJECTIVES
Distinguish completed observations from recruitment workload, document sample assumptions, and plan staffing and logistics without quietly changing the study design.
THE CALCULATOR IS THE BEGINNING OF OPERATIONS
A sample target is useful only when it is linked to the population, selection stages, eligible-person rule, expected response and analysis plan. Review the approved or proposed design before translating the number into households, contacts, interviews or staff-days.
For a single-prevalence objective, specify expected prevalence, absolute precision, confidence, design effect and response assumptions. A two-group comparison, trial or other primary objective needs an appropriate alternative calculation. Small subgroups may need separate planning. Do not imply that a single total guarantees all secondary analyses.
WORKED FICTIONAL PLAN
Use the LMS workshop with prevalence 50%, precision ±5 percentage points, confidence 95%, design effect 1 and response 90%, leaving population size blank. The recruitment target is 427. With design effect 1.5 and the other inputs fixed, it is 641. Round up at the final step rather than repeatedly inflating rounded intermediate values.
These targets count the eligible units specified in the plan. If the unit is a child, the target is not automatically a household target. Households may have no eligible child, several eligible children or an unavailable caregiver. Separate selection, eligibility and response assumptions.
Suppose the team expects about 18 completed interviews per team-day, after pilot testing. Estimating staff-days from the expected completed sample can help budgeting, but travel, listing, repeat contacts, supervision, training and data reconciliation add work. Do not raise the assumed productivity merely to fit a predetermined deadline.
BUILD A WORKLOAD REGISTER
List selection/listing tasks, training, translations, equipment, transport, appointments, interviews, repeat visits, supervision and daily data checks. Assign responsibilities, dependencies and contingency resources. Separate confirmed capacity from assumptions.
A budget-driven revision to sampling, precision or eligibility is a methodological change. Document it, assess the effects on the objectives and obtain the appropriate review before field implementation. Do not solve workload problems through convenient substitution of nearby households.
YOUR TASK
Reproduce the two targets in Research lab → Sample size, saving every input. Create a fieldwork plan that converts your selected target into contact, interview, travel and quality-check tasks. Add three risks—such as inaccessible areas, low response or equipment failure—with a response and decision owner.
https://aces-af.org/learn/index.php?r=tool&tool=sample
SELF-CHECK AND FEEDBACK
Is dividing by 0.90 identical to adding 10%? No. Is a larger assumed design effect a complete plan for cluster sampling? No; cluster numbers, size, selection and analysis still need specification. Is an estimated team-day a guarantee? No; validate productivity in a pilot and retain contingencies.
PORTFOLIO OUTPUT
A justified target, an assumptions sheet and a workload/budget outline linked to the sampling plan.
FURTHER READING
WHO STEPS Manual, Part 2: Planning and set-up.
https://www.who.int/docs/default-source/ncds/ncd-surveillance/steps/part2.pdf
Distinguish completed observations from recruitment workload, document sample assumptions, and plan staffing and logistics without quietly changing the study design.
THE CALCULATOR IS THE BEGINNING OF OPERATIONS
A sample target is useful only when it is linked to the population, selection stages, eligible-person rule, expected response and analysis plan. Review the approved or proposed design before translating the number into households, contacts, interviews or staff-days.
For a single-prevalence objective, specify expected prevalence, absolute precision, confidence, design effect and response assumptions. A two-group comparison, trial or other primary objective needs an appropriate alternative calculation. Small subgroups may need separate planning. Do not imply that a single total guarantees all secondary analyses.
WORKED FICTIONAL PLAN
Use the LMS workshop with prevalence 50%, precision ±5 percentage points, confidence 95%, design effect 1 and response 90%, leaving population size blank. The recruitment target is 427. With design effect 1.5 and the other inputs fixed, it is 641. Round up at the final step rather than repeatedly inflating rounded intermediate values.
These targets count the eligible units specified in the plan. If the unit is a child, the target is not automatically a household target. Households may have no eligible child, several eligible children or an unavailable caregiver. Separate selection, eligibility and response assumptions.
Suppose the team expects about 18 completed interviews per team-day, after pilot testing. Estimating staff-days from the expected completed sample can help budgeting, but travel, listing, repeat contacts, supervision, training and data reconciliation add work. Do not raise the assumed productivity merely to fit a predetermined deadline.
BUILD A WORKLOAD REGISTER
List selection/listing tasks, training, translations, equipment, transport, appointments, interviews, repeat visits, supervision and daily data checks. Assign responsibilities, dependencies and contingency resources. Separate confirmed capacity from assumptions.
A budget-driven revision to sampling, precision or eligibility is a methodological change. Document it, assess the effects on the objectives and obtain the appropriate review before field implementation. Do not solve workload problems through convenient substitution of nearby households.
YOUR TASK
Reproduce the two targets in Research lab → Sample size, saving every input. Create a fieldwork plan that converts your selected target into contact, interview, travel and quality-check tasks. Add three risks—such as inaccessible areas, low response or equipment failure—with a response and decision owner.
https://aces-af.org/learn/index.php?r=tool&tool=sample
SELF-CHECK AND FEEDBACK
Is dividing by 0.90 identical to adding 10%? No. Is a larger assumed design effect a complete plan for cluster sampling? No; cluster numbers, size, selection and analysis still need specification. Is an estimated team-day a guarantee? No; validate productivity in a pilot and retain contingencies.
PORTFOLIO OUTPUT
A justified target, an assumptions sheet and a workload/budget outline linked to the sampling plan.
FURTHER READING
WHO STEPS Manual, Part 2: Planning and set-up.
https://www.who.int/docs/default-source/ncds/ncd-surveillance/steps/part2.pdf
My study notes
Stored only in this browser. Do not enter patient information.Saved in this browser. No account needed.