Application of Structural Equation Models to Estimate One Personality Trait From Others With Parental Mental Health as a Latent Variable

Abstract

Introduction: Adolescence is a transformational but vulnerable age.Personality starts taking shape at this age and the traits developed during this phase are likely to remain with a person for the whole life. Afavourable atmosphere at home and surroundings play a pivotal role in development in different dimensions of personality. Lack of a support system may result in some psychiatric disorders among adolescents which might needs to be addressed. Eysenck personality questionnaire is a psychiatric tool to assess various dimensions of personality. Methods: Structural equation models with two and three layers were applied to estimate personality traits as identied by Eysenck personality questionnaire on two groups of adolescents: (i) The 'Control' group whose parents were not reported any psychiatric disorder; and (ii) the 'Case' group whose parent(s) were suffering from some psychiatric disorder as diagnosed by DSM-IV. Results: The conventional three-layer model for 'Control' group estimated 'Psychoticism' directly by 'Lie-scale' and indirectly by 'Extraversion' with p-values < 0.05. The two-layered model yielded no signicant difference between the base and the proposed model for 'Control' and 'Case' group as p-values were more than 0.05. The variability of Neuroticism was estimated up to 63.6% by a linear combination of Psychoticism, Lie-scale and Gender in Layer 1 of the model for the 'Control' group. Conclusion: In this study either through the conventional three-layer model or the proposed two-layer models, we concluded that some information about one dimension can be extracted from the other dimension with EPQ even if the correlation is not signicant. Also, the proposed two-layer model established the effect of gender at 10% level of signicance in case of 'Control' group. Psychoticism and Neuroticism mean scores were higher for the 'Case' groups than those for the 'Control' group thus establishing the effect of parental mental health on adolescents' personality