CACE as a policy-relevant parameter for intervention effects on students receiving a meaningful dose of treatment¶

Theory · All theories
Evidence · no claims cited

Description¶

The article frames the complier average causal effect (CACE) parameter as the estimand in education RCTs that pertains to students who actually receive a meaningful dose of treatment services. It is presented as "often of policy interest" because it targets intervention effects for the treated complier subgroup rather than all assigned students. The article develops this parameter within a causal inference and instrumental variables framework for two-level clustered designs.

Design Implications¶

Context¶

Requirements¶

• Applicable in randomized control trials where treatment receipt differs from assignment, so compliers can be defined

Constraints¶

• Framed for two-level clustered RCTs in the education field

Target Learners¶

• Students in education RCTs

Target Learning Objectives¶

• Estimating intervention effects for students receiving a meaningful dose of treatment services

Claims¶

Related Theories¶

Examples¶

• Ten-RCT empirical comparison of CACE variance formulas
• Use CACE analysis alongside ITT to estimate effects of teacher coaching accounting for implementation compliance

Key Sources¶

• Peter Z. Schochet, Hanley S. Chiang. (2011). Estimation and Identification of the Complier Average Causal Effect Parameter in Education RCTs. Journal of Educational and Behavioral Statistics, vol. 36, no. 3. https://eric.ed.gov/?q=Estimation and Identification of the Complier Average Causal Effect Parameter in Education RCTs