The Three E’s of Quantitative Research: Estimand, Estimator, Estimate

Term	One-line definition	Where it “lives”	Simple example (blood-pressure drug)
Estimand	The scientific question expressed as a precise quantity you want to know.	Before data are collected.	“The average reduction in systolic BP (mm Hg) if every adult took Drug X instead of placebo for 12 weeks.”
Estimator	The recipe or formula you will apply to data to target the estimand.	In your analysis plan / code.	“Difference in sample means between the Drug X group and placebo group (unadjusted).”
Estimate	The numerical answer you get after running the estimator on the actual dataset.	In your results table.	“–7.4 mm Hg (95 % CI –9.1 to –5.7).”

Start with the estimand.
- Clarifies what effect, in whom, and under which conditions you care about.
Choose an estimator that is valid if the required causal/statistical assumptions hold.
- Could be as simple as a risk difference, or as complex as a targeted-maximum-likelihood algorithm.
Compute the estimate once you have data.
- This single number (plus its uncertainty) is what you report.

Estimand ≠ Estimator: The former is a target; the latter is a tool.
Different estimators (e.g., regression with covariate adjustment, IPW, G-computation) can aim at the same estimand.
The estimate is data-dependent and will change with a new sample; the estimand and estimator do not.