EP1: The Core Trio: Estimand, Estimator, and Estimate
Listen to This Episode
Want to dive deeper into the causal trio? This podcast episode provides more in-depth information and intuition about the estimand, estimator, and estimate. So do tune in!
The Core Trio: Estimand, Estimator, Estimate
At the heart of the estimation process are three key terms that sound similar but have distinct meanings: the estimand, the estimator, and the estimate(Neal 2020, 15â18).
Estimand: This is the âwhatâ. It is the specific quantity or theoretical value you are interested in estimating. Think of it as the precise question you want to answer, like âWhat is the average causal effect of a new drug on patient recovery time?â
Estimator: This is the âhowâ. It is the rule, algorithm, or function that you apply to your data to get your answer. In essence, it is a function that takes your data as input and produces an estimate of your estimand.
Estimate: This is the âresultâ. It is the concrete numberâthe approximation of your estimandâthat you get after applying the estimator to your dataset.
The entire process of starting with a question (the estimand) and using data to get a number (the estimate) is called estimation.
An Analogy for the Aspiring Baker
As someone who loves to bake, this parallel really helped these concepts click for me(Dumas 2023, 10â11).

The image beautifully illustrates the concept: your goal is the perfect cake (the estimand), the recipe you follow is your method (the estimator), and the cake you actually pull out of the oven is the result you get (the estimate).