The Trial · Tamil Nadu Critical Thinking Curriculum · India Institute
An India Institute Project
indiai.org ↗
↑ An India Institute Project
Tamil Nadu Critical
Thinking Curriculum
பள்ளியில் பகுத்தறிவு
Phase 1 · Randomised Controlled Trial
RCT Complete · Report Forthcoming
Chapter 04 · Study Design

The Trial

The design, process, and methodological choices behind the randomised controlled trial: 41 schools, 6,711 students, and a rigorous approach to establishing causation.

Chapter 4 of 6

Why a randomised controlled trial

The central research question was whether teaching critical thinking as a co-curricular subject to government school students produces a measurable improvement in their reasoning ability. To answer this with confidence, the study was designed as a randomised controlled trial (RCT). This is the most rigorous method available for establishing that an intervention caused an observed outcome, rather than simply being associated with it.

Schools were assigned to treatment and control groups through randomisation, ensuring that any difference in end-line outcomes between the two groups could be attributed to the CT sessions rather than to pre-existing differences between schools or students. The trial was conducted with the permission and support of the Tamil Nadu School Education Department.

New to RCT methodology? The Knowledge Base explains what a randomised controlled trial is, why it is considered the gold standard for causal inference, and what makes a result credible. Read the Knowledge Base →

Sample and randomisation

From the pool of eligible schools in the urban education block, 41 were selected and randomised: 19 were assigned to the treatment group and 22 to the control group. Treatment school students received the critical thinking curriculum; control school students continued with their regular timetable and were not exposed to the CT sessions.

41 Schools in the trial 19 treatment · 22 control
6,711 Students assessed Grades 8 & 9
1 Urban education block All schools in the same block

Keeping all schools within a single urban education block was appropriate at the curriculum pilot stage. It reduced variation in administrative context, ensured comparable school infrastructure, and kept implementation manageable. A larger-scale pilot, should it proceed, would extend across multiple blocks with randomisation at multiple levels.

Timeline and three stages

The project ran across three stages over 18 months, from September 2022 to April 2024.

September–October 2022
Stage 1: Preparation and baseline
A baseline proficiency survey was administered to all students in both treatment and control schools, covering Tamil, English, and Mathematics. The purpose was to establish each student's academic starting point, providing control variables for the impact estimation at end-line. Randomisation was completed after the baseline survey. CT tutors were recruited and trained in the curriculum.
The baseline used tools developed by the Annual Status of Education Report (ASER Centre). Each student was tested individually by a teaching assistant in the school premises. The assistant encouraged students to attempt each question more than once and helped them understand what was being asked, without helping with the answer. The rationale was to avoid underestimating students' actual ability: a baseline that records lower scores than students are genuinely capable of would bias the end-line effect estimate upward, making the intervention appear more effective than it is.
November 2022–August 2023
Coordination and timetable integration
Following randomisation, the project team worked with each treatment school to schedule one dedicated CT session per week into their existing timetables. This was a substantial coordination exercise, carried out across multiple school administrations. Integrating a new subject into a live school schedule required sustained engagement with school leaders and teachers, and took several months to finalise across all 19 treatment schools.
September 2023–March 2024
Stage 2: Intervention
Trained CT tutors delivered one-hour sessions per week in treatment schools. Each session followed the structured curriculum described on The Curriculum page. Students in control schools continued their regular timetable throughout this period, with no exposure to the CT sessions. 30 sessions were planned across the intervention period.

The 12 tutors were recruited and trained by the project team; they were not regular school teachers. All were women between the ages of 20 and 45, each holding at minimum an undergraduate degree, with three holding a BEd. Tutors were assigned at the classroom level rather than the school level, and moved between schools to teach their assigned classes according to the agreed timetable. Classroom assignments were randomised where possible; where the travel distance between a tutor's assigned schools made this impractical, assignments were made on a convenience basis.
April 2024
Stage 3: End-line assessment
The end-line assessment was administered to all students in both treatment and control schools, measuring critical thinking and reasoning ability. A student motivation survey was also administered at this stage.

Session delivery

The planned 30 sessions were not fully delivered in all treatment schools. Disruption to the school calendar from unexpected developments during the intervention period reduced the number of sessions that could be conducted. Across treatment schools, sessions delivered ranged from 5 to a maximum of 19.

This variance in dosage was accounted for in the analysis. A per-session effect analysis was conducted alongside the main impact analysis, assessing the relationship between the number of sessions a student received and their end-line score. Results from both analyses will be published with the full report.

End-line assessment: design and administration

The end-line CT assessment used questions presented as scenarios drawn from everyday situations familiar to students: financial decisions, social claims, and common information environments. Some questions were adapted from established assessments, including the 2020 Critical Thinking Test by the Victorian Curriculum and Assessment Authority (Australia) and assessments from the William and Mary School of Education (United States). Others were developed originally by the project team. All were contextualised for Tamil Nadu in language, setting, and cultural register.

The endline assessment was designed to be consistent with the curriculum's oral-first approach. Students should not be disadvantaged by reading ability when being tested on reasoning ability.

The end-line assessment was administered by volunteers from the Tamil Nadu government's Illam Thedi Kalvi programme. To ensure independence, volunteers were not assigned to schools in which they already volunteered, ensuring that neither the CT tutors nor familiar community members could influence student responses.

Invigilator groups of two or three administered the test in each room. One invigilator read the questions aloud to the class while each student also had a printed question booklet. This arrangement maintained the oral-first approach that ran through the curriculum: students who find reading difficult were not put at a disadvantage during assessment. Answers were subsequently evaluated by trained CT teachers.

The hypotheses

Two research hypotheses guided the study.

Hypothesis 1 · Primary
Critical thinking skills improve with the curriculum.
Teaching critical thinking skills to adolescent school students will produce measurable improvements in their ability to evaluate information, distinguish strong arguments from weak reasoning, and draw sound inferences across a range of topics.
Tested at end-line · Results forthcoming
Hypothesis 2 · Secondary
CT training also improves academic performance in regular subjects.
Teaching critical thinking skills will lead to improved academic performance in regular curricular subjects, including mathematics and language, even when CT lessons are taught as a discrete subject rather than integrated into the regular curriculum. Baseline proficiency data in Tamil, English, and Mathematics was collected for this purpose. However, the end-line assessment for regular subjects could not be administered: the announcement of general elections advanced the school year-end examinations, leaving no window in which to conduct independent testing. Results for Hypothesis 2 will therefore not be reported.
Could not be tested · Reason stated above
Student motivation
Effect on student motivation.
In addition to the CT assessment, a student motivation survey was administered at end-line using an instrument adapted from the Young Lives longitudinal study. This component assessed whether participation in the CT programme affected student motivation and engagement with learning more broadly.
Tested at end-line · Results forthcoming

What comes next

The RCT is complete. Data collection is done and analysis has been carried out. The full report, covering results, methodology, effect sizes, and policy implications, is forthcoming. When it is released, all findings will be published here in full.