Pre-Randomization Compliance Screening Filter
RCTClinical Trial
Enrolling subjects into longitudinal mobile field experiments without evaluating early engagement leads to post-randomization attrition and statistical noise. Applying a short pre-randomization compliance filter removes non-active participants while preserving sample comparability.
Picture this
Think of a 30-day trial period at a health club before issuing a full membership; only people who swipe their card at least twice during the trial month are officially enrolled in the long-term fitness tracking program, removing those who signed up but never intended to return.
What the evidence says
Retained 4,458 out of 8,248 initial recruits (54% retention rate) while maintaining statistical balance across observable baseline demographics between retained and non-retained cohorts (e.g., 74% female in retained vs 66% in non-retained).
- Who was studied
- 8,248 initially recruited Filipino migrants in Dubai and Sharjah, screened down to 4,458 retained study participants.
- How
- Mandatory 4-week active monitoring period requiring participation in at least two weekly micro-surveys prior to treatment assignment at weeks 6–9.
What to do
Implement a mandatory multi-week active participation filter prior to experimental randomization to exclude unengaged subjects without introducing selection bias.
From the source
"We therefore considered a study participant to be retained in the sample only if they responded to at least two out of the first four weekly remittance surveys regardless of whether they actually recorded any remittances."
A Field Experiment among Filipino Migrant Workers in the UAE