Related Experiment Video
Updated: Feb 7, 2026

Laboratory Estimation of Net Trophic Transfer Efficiencies of PCB Congeners to Lake Trout Salvelinus namaycush from Its Prey
Published on: August 29, 2014
Bounded, efficient and multiply robust estimation of average treatment effects using instrumental variables.
Linbo Wang1, Eric Tchetgen Tchetgen1
1Harvard T.H. Chan School of Public Health, Boston, Massachusetts, U.S.A.
This study introduces new assumptions for instrumental variables (IVs) to fully identify the average treatment effect (ATE), overcoming limitations of standard IV models. New estimators are developed for robust causal inference, especially for binary outcomes.
Area of Science:
- Econometrics
- Causal Inference
- Statistics
Background:
- Instrumental variables (IVs) are crucial for estimating causal effects amid unmeasured confounding.
- Standard IV models often result in only partial identification of the average treatment effect (ATE).
Purpose of the Study:
- To propose novel assumptions enabling the full identification of the ATE.
- To develop robust estimators for causal effect estimation under various data models.
Main Methods:
- Separation of identification assumptions from estimation model assumptions.
- Construction of multiple and multiply robust estimators.
- Development of bounded estimators for binary outcomes.
Main Results:
- Achieved full identification of the ATE under proposed assumptions.
- Developed estimators consistent across multiple observed data models.
- Obtained bounded ATE estimators (-1 to 1) for binary outcomes.
Conclusions:
- The proposed methods enhance the identification and estimation of causal effects using instrumental variables.
- The approach offers greater flexibility by decoupling identification from specific data models.
- The findings are applicable to various fields, including the analysis of education's effect on earnings.
Related Concept Videos
Uncertainty in Measurement: Reading Instruments
Average Acceleration
Design Example: Capacitance Multiplier Circuit
The circuit illustrated in Figure 1 below incorporates two op-amps, with the first operating as a voltage follower and the second acting as an inverting amplifier.
Average Velocity
Average Value of a Function
Average Power

