UCL Discovery
UCL home » Library Services » Electronic resources » UCL Discovery

A statistical study of the nature of periodontal disease progression

Sterne, Jonathan Abram Cutler; (1990) A statistical study of the nature of periodontal disease progression. Doctoral thesis (Ph.D), UCL (University College London). Green open access

[img]
Preview
Text
A_statistical_study_of_the_nat.pdf

Download (5MB) | Preview

Abstract

The thesis is concerned with questions arising from recent disputes about the nature of periodontal disease progression. Relevant periodontal and statistical literature is reviewed. Three models for periodontal disease progression, allowing for measurement error, are constructed; for (1) constant progression, (2) instantaneous bursts of activity and (3) varying, but non-instantaneous, rates of progression. When the covariance structures of the three models are examined, it becomes clear that they are hierarchical, with (1) being a limiting case of (2) and (2) a limiting case of (3). The covariance structure estimation methods of Browne do not require restrictive assumptions about the underlying distribution. Software using these methods was written in APL. Simulation experiments were performed to examine the conditions under which it is possible to distinguish data from the three models. The study of Löe et al. on Sri Lankan tea labourers is the largest longitudinal study of the natural history of periodontal disease. Preliminary analysis of these data reveals a large subject intraclass correlation, in contrast to recent claims made in the periodontal literature. The parameters and goodness of fit of models (1)-(3) for the data from this study are estimated, together with cross-validation coefficients and model fit indices to aid model selection. A criterion for designing studies of periodontal disease progression might be the minimisation of the asymptotic variance of a parameter which distinguishes models (1) and (2). This is calculated, and the asymptotic properties of maximum-likelihood estimates under the minimising conditions are investigated. Possibilities for further work are briefly discussed. It is to be hoped that the methods will be applied to other data collected in longitudinal studies of periodontal disease progression.

Type: Thesis (Doctoral)
Qualification: Ph.D
Title: A statistical study of the nature of periodontal disease progression
Open access status: An open access version is available from UCL Discovery
Language: English
Additional information: Thesis digitised by ProQuest.
Keywords: Biological sciences; Health and environmental sciences; Periodontal disease
URI: https://discovery.ucl.ac.uk/id/eprint/10108218
Downloads since deposit
11Downloads
Download activity - last month
Download activity - last 12 months
Downloads by country - last 12 months

Archive Staff Only

View Item View Item