Building Resilience through Emotional Self Regulation in Wartime Ukraine

Last registered on September 21, 2026
View Trial History

Pre-Trial

Trial Information

General Information

Title
Building Resilience through Emotional Self Regulation in Wartime Ukraine
RCT ID
AEARCTR-0019721
Initial registration date
September 15, 2026

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
September 21, 2026, 10:10 AM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Country
Ukraine
Region
Kyiv

Primary Investigator

Name
Monika Pompeo
Affiliation
Kyiv School of Economics
Contact Primary Investigator

Other Primary Investigator(s)

PI Name
Dariia Mykhailyshyna
PI Affiliation
Kyiv School of Economics
Contact Investigator
PI Name
Giuseppe Sorrenti
PI Affiliation
HEC Lausanne
Contact Investigator

Additional Trial Information

Status
On going
Start date
2026-09-14
End date
2027-12-31
Keywords
Behavior, Crime, Violence, & Conflict, Health
Additional Keywords
Education; socio-emotional skills; children; self-regulation; conflict-affected settings
JEL code(s)
JEL I21, I24, C93, D91
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
We evaluate Feelings Detectives, a teacher-delivered programme designed to support emotional self-regulation among primary-school children in Kyiv, Ukraine. The programme teaches children to recognise feelings and body cues, select and use regulation strategies, return to goal-directed activity, seek appropriate help and manage peer interactions. It comprises ten weekly classroom lessons with supporting routines and materials.

The study uses a classroom-randomised waitlist-control design within seven schools. The confirmatory sample comprises 52 classrooms in Grades 2-4, with 26 assigned to immediate programme delivery and 26 to waitlist control. Four Grade 1 classrooms are randomised separately and will be analysed only as exploratory. Outcomes are measured using tablet-based behavioural tasks, child reports, teacher ratings and peer nominations.

Data are collected from both groups at three points: baseline; a first endline after the initial treatment group completes the programme, and a second endline, moths after the treatment group has finished. The first endline identifies the effect of programme assignment relative to the not-yet-treated control group. The second endline assesses whether outcomes persist several months after programme completion.

The study examines whether assignment to Feelings Detectives improves children’s emotional understanding, inhibitory control and stability of incentivised choices. Four primary endpoints capture these domains: taught-cue and transfer-item emotion-recognition accuracy, Hearts and Flowers mixed-block accuracy, and switching across identical time and risk choices administered before and after a standardised classroom activity. Secondary outcomes assess teacher-reported regulation, social behaviour and peer relationships, while exploratory outcomes include wellbeing and academic task performance. Classroom randomisation identifies the effects of programme assignment relative to waitlist control. Changes in choice stability will be interpreted alongside the other outcomes, without assuming that every switch reflects a failure of self-regulation or that the within-session contrast isolates the effect of emotional activation.
External Link(s)

Registration Citation

Citation
Mykhailyshyna, Dariia, Monika Pompeo and Giuseppe Sorrenti. 2026. "Building Resilience through Emotional Self Regulation in Wartime Ukraine." AEA RCT Registry. September 21. https://doi.org/10.1257/rct.19721-1.0
Sponsors & Partners
Experimental Details

Interventions

Intervention(s)
Feelings Detectives is a teacher-delivered programme supporting emotional understanding and self-regulation among primary-school children in Kyiv, Ukraine. It comprises ten approximately weekly, 40-minute classroom lessons, supported by teacher training, teaching materials and short classroom routines. Activities address recognising emotions and bodily cues, understanding emotional intensity, using regulation strategies, responding to frustration, returning to activities, seeking help and managing peer interactions. Treatment classrooms receive the programme during the evaluation period. Waitlist classrooms continue their usual schooling and are scheduled to receive the programme after the main endline assessment.
Intervention Start Date
2026-09-14
Intervention End Date
2027-12-31

Primary Outcomes

Primary Outcomes (end points)
The primary outcomes, measured at the first post-programme endline before waitlist delivery, cover three domains:
1. Emotional understanding: taught-cue emotion-recognition accuracy and transfer-item emotion-recognition accuracy.
2. Inhibitory control: Hearts and Flowers mixed-block accuracy.
3. Choice stability: total item-level switching across repeated incentivised time and risk choices administered before and after a standardised classroom activity.
Primary Outcomes (explanation)
Emotion recognition produces two separate scores: the number of correct responses to four taught-cue items, ranging from 0–4, and two transfer items, ranging from 0–2. Higher scores indicate greater accuracy. The items and scoring remain unchanged across assessment waves.

Hearts and Flowers accuracy is the proportion of correct responses across 33 scored mixed-block trials. Practice trials are excluded. Incorrect responses and recorded response-window timeouts count as incorrect; trials missing because an assessment was interrupted are not automatically coded as errors.

Choice switching is the number of decisions that differ between two administrations of identical choice lists: four time choices and four risk choices, separated by a researcher-led Simon Says activity. The score ranges from 0–8; lower scores indicate greater stability, not necessarily better preferences or definitive evidence of emotional regulation. All eight matched choice pairs are required to construct this endpoint.

These are four separate primary endpoints, not a combined index. We will apply Holm correction across the four primary tests, controlling the family-wise error rate at 5 percent, and report all estimates.

Secondary Outcomes

Secondary Outcomes (end points)
Secondary outcomes comprise three families: teacher-reported child regulation and classroom behaviour; social behaviour and peer relationships; and time and risk choices. Exploratory outcomes include child-reported wellbeing, distress, regulation self-efficacy, future expectations, academic task performance and implementation measures.
Secondary Outcomes (explanation)
Teacher outcomes are eight separate ratings of stopping when calm, stopping when emotionally activated, awareness of emotional intensity, independent strategy use, recovery, persistence, regulation during conflict and helping peers. We also examine the activated-minus-calm stopping contrast. Ratings use a five-category scale; insufficient opportunity to observe is treated as missing. No overall eight-item score is assumed.
Social outcomes include stars sent in the trust game, the mean fraction returned across four contingent receipt amounts, stars allocated in the giving game, and isolation in friendship and help networks. The return measure is a descriptive task summary, not a validated scale. Network isolation means receiving no nomination of the relevant type among participating nominators; unavailable nomination data are not coded as isolation.
Choice outcomes include time and risk switching separately and the numbers of delayed-reward and risky choices in the first administration. Greater patience, risk-taking or giving will not automatically be interpreted as improvement.
The exploratory distress measure sums four study-specific seven-day items, ranging from 0–12 when all responses are observed. It is not a validated PHQ-4 administration and will not use clinical cutoffs. Secondary tests will use Holm correction within each of the three named families; exploratory analyses will be identified separately.

Experimental Design

Experimental Design
The study is a classroom-randomised controlled trial in seven Kyiv schools. The confirmatory sample consists of 52 classrooms in Grades 2–4, with 26 assigned to the programme and 26 to waitlist control. Four Grade 1 classrooms are randomised separately and analysed as an exploratory sample. Measurements include tablet-based child assessments, teacher reports and peer nominations. The main endline is scheduled before waitlist programme delivery. Effects will be estimated by original classroom assignment, regardless of programme attendance.
Experimental Design Details
Not available
Randomization Method
Computer-generated restricted randomisation in Stata. The confirmatory draw used seed 13082026 and selected uniformly from a retained pool of balanced eligible allocations. The separate Grade 1 draw used seed 13082027.
Randomization Unit
Classroom. Grades 2–4 form the confirmatory sample; Grade 1 classrooms are randomised separately for exploratory analysis.
Was the treatment clustered?
Yes

Experiment Characteristics

Sample size: planned number of clusters
56 classrooms across seven schools: 52 confirmatory classrooms in Grades 2–4 and four exploratory Grade 1 classrooms. Given the nature and complication of the experiment, some drop outs might be expected.
Sample size: planned number of observations
1,157 pupils on the assignment roster: 1,078 in the confirmatory sample and 79 in the exploratory Grade 1 sample. Actual assessed numbers depend on consent, attendance and follow-up.
Sample size (or number of clusters) by treatment arms
Confirmatory sample: 26 treatment classrooms with 531 rostered pupils; 26 waitlist classrooms with 547 rostered pupils. Exploratory Grade 1 sample: two treatment classrooms with 39 pupils; two waitlist classrooms with 40 pupils. Overall: 28 treatment classrooms with 570 pupils and 28 waitlist classrooms with 587 pupils.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Preliminary normal-approximation calculations use 52 confirmatory classrooms, 1,078 pupils, equal classroom sizes, no attrition and no baseline-outcome adjustment. With 80 percent power and a conservative two-sided significance threshold of 0.0125 for four primary endpoints, the approximate MDE is 0.29 outcome standard deviations at an ICC of 0.05, 0.35 at an ICC of 0.10, and 0.45 at an ICC of 0.20. These are illustrative sensitivities, not final outcome-specific calculations; they do not incorporate unequal classroom sizes, finite-cluster inference or the exact restricted randomisation. Percentage-point equivalents require outcome-specific variances
Supporting Documents and Materials
IRB

Institutional Review Boards (IRBs)

IRB Name
Research Ethics Committee of the Kyiv School of Economics
IRB Approval Date
2026-08-13
IRB Approval Number
514B5C86A1E5DA1104000000831F340004DD6605
Analysis Plan