General Mental Health in Adolescence: Conceptualisation and Measurement Issues

Abstract

Measuring adolescent mental health in general population samples is vital to estimating prevalence, understanding risk, and assessing intervention. Clinically rated methods suffer from ontological and reliability problems, and parent and teacher ratings typically have poor convergence with young people’s self-reports. Self-reports also provide direct access to thoughts and feelings, and there are increasing calls to hear directly from young people. Despite a clear need for robust approaches to self-reported mental health in adolescence, psychometric development standards have tended to be poor. The current thesis aimed to provide insight into more robust approaches to measuring general mental health via four papers, including secondary analysis of various datasets from the HeadStart project. Paper 1 considered the construct-level relationship between symptom and wellbeing domains. Internalizing symptoms showed equally strong relationships to wellbeing and externalizing problems, and there was also evidence for a general internalizing distress factor. Paper 2 explored indicator-level interactions over three years between internalizing symptoms, wellbeing indicators and psychosocial correlates via a multiverse framework. The multiverse framework demonstrated that the importance of indicators in the network was often sensitive to particular item operationalizations, though a few key indicators were consistently important. Paper 3 examined the age appropriateness of the commonly used self-report Strengths and Difficulties Questionnaire. Items were generally found to have inappropriate reading ages and be of low quality, while measurement invariance analysis suggested the measure functioned comparably across younger and older adolescents. Paper 4 is a meta-review of self-report general mental health measures. Content and psychometrics were analyzed. A relatively narrow range of indicators was found across constructs within mental health but measures were generally not interchangeable and had low psychometric quality. The current thesis represents a major step forward for the theoretical and empirical understanding of general mental health. Findings suggest emotionally-focused indicators, including happiness and worry, could be particularly important when examining prevalence, risk factors, or intervention response, though more work is needed to confirm indicators with young people. The thesis also highlights a broader need to develop new measures.

Date of Award	31 Dec 2022
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Margarita Panayiotou (Supervisor) & Neil Humphrey (Supervisor)