Active Tracking Attrition Mitigation

In Simple Words

Dispatching field teams directly to children's homes when they miss school testing operates like a home delivery service tracking down uncollected mail to ensure every package is accounted for.

Research question

High rates of student absenteeism, urban migration, and local civil unrest in developing contexts lead to panel attrition that can severely bias impact evaluation estimates.

Participants

N = 12,855 (Year 1) and N = 21,936 (Year 2) primary students in Mumbai and Vadodara, India (including tracking during major civil unrest in 2002).

Study Design

Active field tracking protocol utilizing repeat school visits and direct home testing of absent students.

Results

Attrition was maintained at 3.4% to 7.7% across most study arms (except Vadodara Year 1 at 17-18% due to communal riots), with attritor pretest scores showing no significant difference between treatment and control groups.

Takeaway

Deploy direct home-visit testing protocols for absent participants to minimize differential attrition bias in field experiments.

Related findings

• Cross-Sample Intervention Replicability →Methods & evidence

  Imagine testing a new training routine on one sports team and seeing players run faster; to prove the routine actually works and wasn't just luck with that specific team, a coach applies the exact same routine to a second, completely separate team in a different league and verifies that their speed improves by the same margin.

• Hawthorne Effect Mitigation in Phone Surveys →Methods & evidence

  Imagine keeping a daily food diary—just the act of writing down every meal might make a person eat healthier even if they are not on an official diet. Standardizing survey phrasing to mirror baseline surveys ensures that frequent calls act purely as an invisible measuring tape without nudging participants to change their actual habits.

• High-Frequency Survey Recall Validation →Methods & evidence

  Imagine testing whether an accounting class helps people manage money by asking them to estimate their monthly spending at the end of thirty days. To double-check if their end-of-month estimates are accurate or just wild guesses, a researcher checks in with them every single evening to log every penny spent that day, comparing the daily total against the monthly memory estimate.

• Non-Response Bounding and Weighting Robustness →Methods & evidence

  Think of a survey where some people drop out before answering questions. To test whether missing responses distort the final score, researchers test the worst-case scenario by assuming everyone who dropped out got the lowest possible score, and then assuming they all got the highest possible score.