TY - BOOK

T1 - A study into drug-trying behaviour among young people in England

T2 - categorical analysis models in the Presence of missing data

AU - Chan, Henry

PY - 2019

Y1 - 2019

N2 - This research reviewed the "Smoking, Drinking and Drug Use among YoungPeople in England" 2010 survey (the Year 2010 Survey) study in terms of itsdata collection, processing and analysis. The research aim was to gain increasedunderstanding of young people’s drug-trying behaviour in England throughappropriate handling of missing data, as well as, to build upon the previouswork done, developing and applying statistical methodologies for analysis ofmultivariate categorical data collected by the Year 2010 Survey study.The main work done in this research included: (1) modifying the original dataset to arrive the useful working data set; (2) conducting exploratory data analysiswith the working data set to identify direction for further empirical investigation;(3) properly handling the missing data problem in the working data setand (4) developing and applying advanced statistical methodologies to furtheranalyse the working data set.Apart from supporting the main findings of the Year 2010 Survey study thatsmoking, drinking and some drug-related socio-demographic covariates were positively associated with the students’ drug-trying behaviour, additional significant results found by the univariate logistic regression models, log-linearanalysis models, two-parameter item response theory models and latent classanalysis models reported that (1) the 15 drugs were highly and positively associated with each other and each drug exerted different extent of influenceson the students’ drug-trying behaviour and (2) generally, students’ drug-tryingbehaviour could be further explained by numerous smoking, drinking and drug related socio-demographic factors at different extent.These additional findings contributed to a deeper understanding of the drug useproblem, added evidence to the drug related research literature and providedhelpful guidance on formulating policies to combat against drug use problem inEngland. Another contribution of this research was the development of a newmethodology for backward elimination of latent class analysis models whichprovided a more thorough evaluation of the optimal number of latent class andcovariate elimination from saturated model.

AB - This research reviewed the "Smoking, Drinking and Drug Use among YoungPeople in England" 2010 survey (the Year 2010 Survey) study in terms of itsdata collection, processing and analysis. The research aim was to gain increasedunderstanding of young people’s drug-trying behaviour in England throughappropriate handling of missing data, as well as, to build upon the previouswork done, developing and applying statistical methodologies for analysis ofmultivariate categorical data collected by the Year 2010 Survey study.The main work done in this research included: (1) modifying the original dataset to arrive the useful working data set; (2) conducting exploratory data analysiswith the working data set to identify direction for further empirical investigation;(3) properly handling the missing data problem in the working data setand (4) developing and applying advanced statistical methodologies to furtheranalyse the working data set.Apart from supporting the main findings of the Year 2010 Survey study thatsmoking, drinking and some drug-related socio-demographic covariates were positively associated with the students’ drug-trying behaviour, additional significant results found by the univariate logistic regression models, log-linearanalysis models, two-parameter item response theory models and latent classanalysis models reported that (1) the 15 drugs were highly and positively associated with each other and each drug exerted different extent of influenceson the students’ drug-trying behaviour and (2) generally, students’ drug-tryingbehaviour could be further explained by numerous smoking, drinking and drug related socio-demographic factors at different extent.These additional findings contributed to a deeper understanding of the drug useproblem, added evidence to the drug related research literature and providedhelpful guidance on formulating policies to combat against drug use problem inEngland. Another contribution of this research was the development of a newmethodology for backward elimination of latent class analysis models whichprovided a more thorough evaluation of the optimal number of latent class andcovariate elimination from saturated model.

U2 - 10.17635/lancaster/thesis/766

DO - 10.17635/lancaster/thesis/766

M3 - Doctoral Thesis

PB - Lancaster University

ER -