Model selection techniques for the covariance matrix for incomplete longitudinal data.

J J Grady; R W Helms

doi:10.1002/sim.4780141302

Home/Archive/PubMed

PubMed · 7481180

Model selection techniques for the covariance matrix for incomplete longitudinal data.

J J Grady·R W Helms

View original source ↗Export metadata ↓

Abstract

In longitudinal studies with incomplete data, where the number of time points can become numerous, it is often advantageous to model the covariance matrix. We describe several covariance models (for example, mixed models, compound symmetry, AR(1)-type models, and combination models) that offer parsimonious alternatives to unstructured sigma. We evaluate each covariance model with longitudinal data concerning cholesterol as the repeated outcome measure. We discuss strategies for deciding the 'best' model and show a graphical technique for judging goodness-of-fit of covariance models.

Explore related subjects

Cholesterol Data Interpretation, Statistical Humans Likelihood Functions Linear Models Longitudinal Studies Male Models, Statistical Multivariate Analysis Journal Article Review

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

J J Grady, R W Helms. 1995-07-15. Model selection techniques for the covariance matrix for incomplete longitudinal data.. https://doi.org/10.1002/sim.4780141302

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related citations

PubMed2007

Modelling effectiveness of internally heterogeneous organizations in the presence of survey non-response: an application to the ULTRA study.

Levels of turbulence and uncertainty in family medicine practices participating in the ULTRA study are measured via a survey from heterogeneous viewpoints within each practice, including clinicians, nursing staff and office staff. In order to examine the effect of the average practice 'chaos' score on screening provided to patients, the analysis needed to account for large amounts of missing survey responses. Patterns of missing data included portions of missing responses within a job group (e.g. one out of three clinicians) to missing responses from whole subgroups within a practice (e.g. all three clinicians) to a whole practice of missing survey responses (e.g. all clinicians, nursing staff and office staff). The expectation-maximization (EM) algorithm, using a hierarchical model for covariate information, is used to handle the missing data. It was found that the varied patterns of missingness of survey responses among the heterogeneous subgroups within the practices had large effects on estimated effects of practice chaos on patient care.

Cholesterol↗

PubMed2007

A model system to study the insertion of cholesterol into a phospholipid monolayer.

Colloidal probe atomic force microscopy (AFM) was used to study the interaction between a surface bearing tethered cholesterol groups and an egg phosphatidylcholine (egg-PC) monolayer. The cholesterol bearing surface was comprised of a mixed self-assembled monolayer comprised of O-cholesteryl N-(8'-mecapto-3',6'-dioxaoctyl)carbamate (CPEO3) molecules and beta-mercaptoethanol formed on a 20 mum diameter gold-coated silica particle. The egg-PC monolayer was adsorbed onto an octadecylthiol monolayer formed on template-stripped gold. The force between the surfaces, as a function of separation, was measured for surface concentrations of CPEO3 from 0 to 100 mol %. At all concentrations there was a long-range repulsive double-layer force due to weak surface charges. At surface concentrations of CPEO3 from 1 to 29 mol % the interaction on the approach of the surfaces showed a maximum in the repulsive force, followed by a small (2-5 nm) jump into a force minimum corresponding to adhesion of the surfaces. On separation, a normalized pull-off force of 1.0-1.6 mN m(-1) was measured. Over the same concentration range, the calculated interaction energy per CPEO3 molecule decreased from 1.1 +/- 0.2 kT to 0.04 kT. At surface concentrations of 35 mol % and above there was no reproducible adhesion between the cholesterol-bearing surface and the phospholipid monolayer. We attribute the occurrence of short-range attraction and adhesion in the 1-29 mol % regime to the insertion of (some) cholesterol groups into the phospholipid monolayer. At higher surface concentrations the efficiency of insertion is reduced due to steric effects. We discuss the experimental results in the light of the energetics of the insertion of a cholesterol molecule into a lipid bilayer.

Cholesterol↗

PubMed2007

Sum frequency generation spectroscopic study of the condensation effect of cholesterol on a lipid monolayer.

Sum frequency generation (SFG) spectra and surface pressure-molecular area (pi-A) isotherms have been obtained for mixed cholesterol-DPPC monolayers with cholesterol mole fractions, x(chol.), from 0 to 1.0, at the air-water interface, under same conditions, at 22 degrees C. Analysis of the spectra indicated that incorporation of cholesterol into the monolayers at 3 mN m-1 greatly increases the conformational and orientational order of the alkyl chains of DPPC, maximizing these properties at x(chol.)=0.4. Analysis also indicated that order in the mixed monolayers at 15 and 35 mN m-1 is not affected by incorporation of cholesterol. The pi-A isotherms measured at 3 mN m-1 for the mixed monolayer with x(chol.)=0.4 have the largest negative deviation of the molecular area relative to those of ideal mixtures (the so-called "condensation effect" of cholesterol), indicating the most thermodynamically stable state. Comparison of results from SFG spectra and pi-A isotherms explicitly proved that the condensation effect can be interpreted in terms of conformational and orientational ordering of the alkyl chains of DPPC.

Cholesterol↗