Can a Drawing or an Essay Reveal How a Child Is Feeling? Exploring New Ways to Detect Mental Health Risk in Schools

ABSTRACT

Background and Purpose

Most mental disorders begin before the age of 18, but schools still rely on questionnaires and on teachers noticing that something is wrong. Students who are quietly struggling are often missed until the problem becomes serious, and fewer than one in ten adolescents who need help receive it.

This project sits within a funded RGC General Research Fund study asking whether ordinary schoolwork — a short questionnaire, a written essay, a drawing — could help identify students at risk earlier and less intrusively than current methods. Clinical diagnostic interviews serve as the reference standard.

Eureka students join for one semester and work on a focused question using data that has already been collected in a completed pilot study. The scope is deliberately small enough to be finished properly within the semester, rather than a fragment of a larger project.

Research Questions (one will be chosen with the student)

1. Who does current screening miss? Compare what a standard questionnaire detects against what a clinical interview detects, and examine whether the gap differs by gender, urban or rural background, or family circumstances.
2. Is this approach workable in real classrooms? Examine how many students completed the essay and drawing tasks, who did not, how long it took, and what this implies for schools considering something similar.
3. What has been tried elsewhere? A structured review of how technology has been used for mental health screening in schools worldwide, what worked, and what was reported poorly.

What students can expect

Students are not expected to arrive with technical skills. Basic data analysis in R or Python is taught at the start, with worked examples, a prepared dataset, and weekly help from the project’s research assistant. Students work only with fully anonymised data. Supervision is by the Principal Investigator, with a research assistant and postgraduate students available for hands-on support, and students are welcome at the project’s regular team meetings.

The deliverable is a short written report and a presentation. Where the work is strong, students may be supported to develop it into a conference abstract or a publication.

SKILLS & EXPERIENCE REQUIRED

• Open to students from any discipline; public health, medicine, psychology, statistics, data science and computer science are all a natural fit
• No prior research experience required
• No programming experience required — R or Python is taught at the start of the project
• Comfortable with secondary-school level statistics (averages, percentages, simple comparisons)
• Genuine interest in adolescent mental health or in how health evidence is produced
• Willing to read academic papers in English
• Careful and patient with detail, since real-world data is messy and incomplete
• Able to attend biweekly supervision meetings in person and come prepared
• Willing to sign a confidentiality undertaking and follow data-handling rules strictly
• Comfortable asking questions when stuck, rather than working around a problem alone