# Idea
An identification strategy is an idea or approach that lets researchers use observable quantities (e.g., sample averages) to reveal parameters of interest (e.g., [[average treatment effect]]).
It's the approach you take to approximate a [[randomized controlled trial]] or experiment with just observational data.
# References
- [[Gerber 2012 field experiments]]
- [[Angrist 2008 mostly harmless econometrics]]