08. Defend a complete design and sampling dossier
Lesson
LEARNING OBJECTIVES
Integrate the design brief, evidence gap, timeline, causal assumptions, bias register and sampling plan, then revise the proposal in response to structured critique.
YOUR DESIGN DEFENSE
Use the fictional missed-appointment or household-water scenario, or a carefully bounded topic of your own. Prepare a concise dossier containing: the decision and uncertainty; primary objective and target quantity; evidence-gap table; comparison of candidate designs; chosen recruitment and timeline; measurement definitions; causal assumptions; sampling and sample-size approach; prevention of important biases; analysis implications; and unresolved ethical or feasibility questions.
Do not add detail merely to make the document longer. Every decision should answer a reader's practical question about who is included, what is measured, when it happens and what conclusions could follow.
WORKED REVIEW COMMENT
“Your target population is all district children, but the frame contains only clinic attendees. Your objective concerns new illness, but you measure only symptoms recalled at a single interview.”
A useful response acknowledges both mismatches. Options might include building a credible household frame for the district population, narrowing the target population explicitly, changing to a prevalence objective, or developing an appropriate follow-up design. The best revision depends on the real decision, resources and ethical considerations. Increasing the sample size alone addresses neither mismatch.
CHECK THE SAMPLING DETAIL
Specify how households are identified, how eligible people are enumerated, how one person is randomly selected if relevant, how many contact attempts are planned and how response is documented. State how selection probabilities and design variables will reach the analysis dataset. Include the source and rationale of each sample-size input rather than only a calculator screenshot.
REVIEW FOR INTERNAL CONSISTENCY
Can another researcher identify the same eligible population from your rules? Are exposure classification and follow-up aligned? Does the outcome definition match the objective? Are proposed adjustment variables justified? Does the analysis respect clustering, matching or unequal probabilities? Is a causal claim being made where only a descriptive association is supportable?
YOUR TASK
Ask a peer or supervisor to review the dossier, or use the questions above for self-review. Record at least three comments in a revision log: issue, significance, proposed change, responsible person and status. Save a dated revised version and a list of decisions requiring specialist input.
Use the Research question builder, Study design explorer and Protocol workshop to organize drafts; the tools do not approve the design or perform complex-survey analysis.
https://aces-af.org/learn/index.php?r=tool&tool=picot
https://aces-af.org/learn/index.php?r=tool&tool=design
https://aces-af.org/learn/index.php?r=tool&tool=protocol
SELF-CHECK AND FEEDBACK
An internally complete dossier may still be scientifically inadequate. Revisit weak assumptions, inaccessible populations or unreliable measures. Acknowledge unresolved issues rather than inventing approvals, data access or local estimates. Successful completion means you can justify and critique the plan, not that real data collection is authorized.
PORTFOLIO OUTPUT
A versioned design dossier and review-response log. Next, use Planning & Conducting Public Health Studies to turn a defensible design into operational field procedures.
FURTHER READING
CDC Field Epidemiology Manual.
https://www.cdc.gov/field-epi-manual/
Integrate the design brief, evidence gap, timeline, causal assumptions, bias register and sampling plan, then revise the proposal in response to structured critique.
YOUR DESIGN DEFENSE
Use the fictional missed-appointment or household-water scenario, or a carefully bounded topic of your own. Prepare a concise dossier containing: the decision and uncertainty; primary objective and target quantity; evidence-gap table; comparison of candidate designs; chosen recruitment and timeline; measurement definitions; causal assumptions; sampling and sample-size approach; prevention of important biases; analysis implications; and unresolved ethical or feasibility questions.
Do not add detail merely to make the document longer. Every decision should answer a reader's practical question about who is included, what is measured, when it happens and what conclusions could follow.
WORKED REVIEW COMMENT
“Your target population is all district children, but the frame contains only clinic attendees. Your objective concerns new illness, but you measure only symptoms recalled at a single interview.”
A useful response acknowledges both mismatches. Options might include building a credible household frame for the district population, narrowing the target population explicitly, changing to a prevalence objective, or developing an appropriate follow-up design. The best revision depends on the real decision, resources and ethical considerations. Increasing the sample size alone addresses neither mismatch.
CHECK THE SAMPLING DETAIL
Specify how households are identified, how eligible people are enumerated, how one person is randomly selected if relevant, how many contact attempts are planned and how response is documented. State how selection probabilities and design variables will reach the analysis dataset. Include the source and rationale of each sample-size input rather than only a calculator screenshot.
REVIEW FOR INTERNAL CONSISTENCY
Can another researcher identify the same eligible population from your rules? Are exposure classification and follow-up aligned? Does the outcome definition match the objective? Are proposed adjustment variables justified? Does the analysis respect clustering, matching or unequal probabilities? Is a causal claim being made where only a descriptive association is supportable?
YOUR TASK
Ask a peer or supervisor to review the dossier, or use the questions above for self-review. Record at least three comments in a revision log: issue, significance, proposed change, responsible person and status. Save a dated revised version and a list of decisions requiring specialist input.
Use the Research question builder, Study design explorer and Protocol workshop to organize drafts; the tools do not approve the design or perform complex-survey analysis.
https://aces-af.org/learn/index.php?r=tool&tool=picot
https://aces-af.org/learn/index.php?r=tool&tool=design
https://aces-af.org/learn/index.php?r=tool&tool=protocol
SELF-CHECK AND FEEDBACK
An internally complete dossier may still be scientifically inadequate. Revisit weak assumptions, inaccessible populations or unreliable measures. Acknowledge unresolved issues rather than inventing approvals, data access or local estimates. Successful completion means you can justify and critique the plan, not that real data collection is authorized.
PORTFOLIO OUTPUT
A versioned design dossier and review-response log. Next, use Planning & Conducting Public Health Studies to turn a defensible design into operational field procedures.
FURTHER READING
CDC Field Epidemiology Manual.
https://www.cdc.gov/field-epi-manual/
My study notes
Stored only in this browser. Do not enter patient information.Saved in this browser. No account needed.