01. From an evidence gap to a precise research objective
Lesson
LEARNING OBJECTIVES
Distinguish a public health problem from a researchable uncertainty, define the quantity you intend to estimate, and align an objective with a realistic decision.
THE DESIGN BRIEF
A problem can be urgent without requiring a new study. First ask what is already known, what is uncertain, who would use the answer and whether available evidence could already support action. Do not claim “no research exists” after reading only a few abstracts.
Separate descriptive, predictive and causal aims. A descriptive study may estimate current service use. A predictive model may forecast future use. A causal study asks how outcomes would differ under specified exposure or intervention strategies. A good model for prediction is not automatically a valid method for estimating a causal effect.
Describe your target quantity in plain language: the population, exposure or intervention contrast if relevant, outcome definition, measurement or follow-up period and comparison measure. This specification is sometimes called an estimand. Decide whether you want a risk, a difference, a ratio, an average measurement or an account of experiences.
WORKED FICTIONAL BRIEF
A district team knows that some caregivers miss child-health appointments. “Improve attendance” is a service goal, not yet a research objective.
Descriptive objective: estimate the proportion of scheduled visits attended among eligible families registered during a defined three-month period.
Exploratory objective: compare recorded attendance between families with different travel times, while describing how travel time and relevant contextual factors are measured.
Intervention objective: estimate the difference in attendance under a specified reminder programme compared with usual scheduling over the same period. This requires a design appropriate to that contrast and explicit assumptions.
These objectives address different uncertainties. They should not share a vague outcome such as “better access.” Define a completed visit, eligibility, cancellations, repeat appointments and the analysis unit before choosing a denominator.
YOUR TASK
Write a six-line brief: decision; existing evidence; remaining uncertainty; primary question; exact target quantity; and the action the findings could reasonably inform. List one secondary objective and label it separately.
SELF-CHECK
Would a percentage of appointments and a percentage of families necessarily be equal? Would an association between travel time and attendance show what a reminder programme would do?
FEEDBACK
No to both. Repeat appointments can make appointment and family denominators differ. An exposure association does not directly identify the effect of a different intervention. Revise the question until the population, unit and comparison match the intended decision.
PORTFOLIO OUTPUT
A design brief with one primary objective and explicit boundaries on interpretation.
SOURCE FOR FURTHER STUDY
CDC Field Epidemiology Manual: Designing and Conducting Analytic Studies in the Field.
https://www.cdc.gov/field-epi-manual/php/chapters/design-conduct-analyze-field-studies.html
Distinguish a public health problem from a researchable uncertainty, define the quantity you intend to estimate, and align an objective with a realistic decision.
THE DESIGN BRIEF
A problem can be urgent without requiring a new study. First ask what is already known, what is uncertain, who would use the answer and whether available evidence could already support action. Do not claim “no research exists” after reading only a few abstracts.
Separate descriptive, predictive and causal aims. A descriptive study may estimate current service use. A predictive model may forecast future use. A causal study asks how outcomes would differ under specified exposure or intervention strategies. A good model for prediction is not automatically a valid method for estimating a causal effect.
Describe your target quantity in plain language: the population, exposure or intervention contrast if relevant, outcome definition, measurement or follow-up period and comparison measure. This specification is sometimes called an estimand. Decide whether you want a risk, a difference, a ratio, an average measurement or an account of experiences.
WORKED FICTIONAL BRIEF
A district team knows that some caregivers miss child-health appointments. “Improve attendance” is a service goal, not yet a research objective.
Descriptive objective: estimate the proportion of scheduled visits attended among eligible families registered during a defined three-month period.
Exploratory objective: compare recorded attendance between families with different travel times, while describing how travel time and relevant contextual factors are measured.
Intervention objective: estimate the difference in attendance under a specified reminder programme compared with usual scheduling over the same period. This requires a design appropriate to that contrast and explicit assumptions.
These objectives address different uncertainties. They should not share a vague outcome such as “better access.” Define a completed visit, eligibility, cancellations, repeat appointments and the analysis unit before choosing a denominator.
YOUR TASK
Write a six-line brief: decision; existing evidence; remaining uncertainty; primary question; exact target quantity; and the action the findings could reasonably inform. List one secondary objective and label it separately.
SELF-CHECK
Would a percentage of appointments and a percentage of families necessarily be equal? Would an association between travel time and attendance show what a reminder programme would do?
FEEDBACK
No to both. Repeat appointments can make appointment and family denominators differ. An exposure association does not directly identify the effect of a different intervention. Revise the question until the population, unit and comparison match the intended decision.
PORTFOLIO OUTPUT
A design brief with one primary objective and explicit boundaries on interpretation.
SOURCE FOR FURTHER STUDY
CDC Field Epidemiology Manual: Designing and Conducting Analytic Studies in the Field.
https://www.cdc.gov/field-epi-manual/php/chapters/design-conduct-analyze-field-studies.html
My study notes
Stored only in this browser. Do not enter patient information.Saved in this browser. No account needed.