Yulia Chekhovska

Yulia Chekhovska

Yulia Chekhovska

Social Impact Analyst working on mental health programme evaluation. Currently at Tokyo English Lifeline (Japan's leading mental health nonprofit).

Available for consulting • Working remotely worldwide

Research Focus

Cost-Effectiveness Analysis

  • DALY-based impact modelling
  • Intervention comparison under uncertainty
  • Sensitivity analysis and assumption testing

Evidence Synthesis

  • Integration of heterogeneous data sources
  • Critical evaluation of empirical studies
  • Working with incomplete or uncertain data

Quantitative Methods

  • Python (data extraction, analysis, NLP)
  • SQL / SOQL (data querying, reporting)
  • Statistical modelling and inference

Domain

  • Mental health and suicide prevention
  • Service delivery and programme evaluation
  • Wellbeing measurement

Background

Current: Conducting cost-effectiveness research on wellbeing interventions and programme evaluations

Education: Research Fellowship in Behavioural Analysis (Chiba University, Japan) | MA German Linguistics (Heinrich Heine University Düsseldorf, Germany) | BA Psycholinguistics (Kyiv National University, Ukraine) | Data Science & AI (Le Wagon)

Technical: Python, SOQL for Salesforce, Excel, basic R

Languages: English (fluent IELTS C1), Ukrainian (native), German (fluent DaF C1), Japanese (JLPT N3/business level)

Research Focus: Mental health intervention effectiveness, wellbeing measurement, willingness to communicate non-parametric quantitative analysis

Published Research

Cost-Effectiveness of Youth Suicide Prevention in Japan

Published on Effective Altruism Forum, 2025

Quantitative evaluation modeling intervention impact using DALYs averted, cost per life saved, and sensitivity analysis. Scraped data from 24+ fragmented government data sources (National Police Agency reports) to build cost-effectiveness estimates for hotline services, school outreach, and community workshops.

Methods: Fermi estimation, data extraction from PDFs using Python, analysed yearly trends and sources of suicide per age group, regional disparities
Key Approach: Modeled three intervention types (hotlines, school programs, workshops) and their impact, calculated cost per DALY under multiple persistence assumptions
Read Full Analysis →

Selected Projects

EA Model

Mental Health Intervention Cost-Effectiveness Model

Outcome: Built a model estimating cost per DALY across three interventions of suicide prevention programme

Methods: Extracted suicide data from 24+ Japanese government PDFs using Python, built structured database, modeled hotline impact, school outreach effectiveness, and workshop outcomes

Tools: Python (data extraction), Excel (modeling), DALYs framework, Fermi estimation

View Details →
NLP Analysis

Programme Evaluation of End-of-Year Curriculum

Outcome: Identified curriculum project satisfaction vs dissatisfaction points across 120+ participant responses

Methods: Applied BERT sentiment analysis to free-text feedback, quantified satisfaction results, recommended programme changes based on findings

Tools: Python, BERT (transformers), statistical analysis

View Details →
Data Analysis

Salesforce Donor Pipeline Management

Outcome: Analysed donation records to identify retention patterns and service effectiveness gaps, moved all data from multiple donor platforms to dashboards for easier reporting to leadership

Methods: Database queries, platform integration, donor retention, App Script for data transfer

Tools: SQL, Salesforce, Excel, Google App Script

Data Pipeline

Health Product Consumer Feedback Data Pipeline

Outcome: Built pipeline collecting YouTube comments for health product satisfaction analysis overtime

Methods: Python API integration, sentiment analysis

Tools: Python, YouTube API, NLP pipelines

View Details →

Get in Touch

Interested in wellbeing research? Want to know more? Let's connect.

LinkedIn GitHub EA Forum