PubMed Health⌕ Search
Home/Archive/PubMed

PubMed · 11213759

Multiple imputation and posterior simulation for multivariate missing data in longitudinal studies.

M Liu·J M Taylor·T R Belin
View original source ↗Export metadata ↓

Abstract

This paper outlines a multiple imputation method for handling missing data in designed longitudinal studies. A random coefficients model is developed to accommodate incomplete multivariate continuous longitudinal data. Multivariate repeated measures are jointly modeled; specifically, an i.i.d. normal model is assumed for time-independent variables and a hierarchical random coefficients model is assumed for time-dependent variables in a regression model conditional on the time-independent variables and time, with heterogeneous error variances across variables and time points. Gibbs sampling is used to draw model parameters and for imputations of missing observations. An application to data from a study of startle reactions illustrates the model. A simulation study compares the multiple imputation procedure to the weighting approach of Robins, Rotnitzky, and Zhao (1995, Journal of the American Statistical Association 90, 106-121) that can be used to address similar data structures.

Explore related subjects

Acoustic StimulationBiometryBlinkingChildElectromyographyHumansLongitudinal StudiesMaleModels, StatisticalMultivariate AnalysisReflex, StartleRegression AnalysisReproducibility of ResultsJournal ArticleResearch Support, U.S. Gov't, P.H.S.

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

M Liu, J M Taylor, T R Belin. 2000. Multiple imputation and posterior simulation for multivariate missing data in longitudinal studies.. https://doi.org/10.1111/j.0006-341x.2000.01157.x

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

KEEP EXPLORING

Related citations

PubMed2002

A universal newborn hearing screening program in Taiwan.

OBJECTIVE: Mackay Memorial Hospital and the Children's Hearing Foundation established a pilot universal newborn hearing screening program in November 1998. Our objective was to assess the feasibility, accuracy and cost effectiveness of implementing universal newborn hearing screening in Taiwan. METHOD: Between November 1998 and October 2000 a total of 6765 newborns were screened for hearing loss prior to discharge from the wellborn nursery at Mackay Memorial Hospital. The average age of the subjects at the initial screening test was 52 h. The program employed a three stage hearing screening protocol using transient evoked otoacoustic emmisions (TEOAE) screening with referral for diagnostic auditory brainstem response assessment. RESULTS: The mean TEOAE screening time per ear was 41.43 s. The overall pass rate at the time of hospital discharge was 93.6%. Thus achieving an acceptable referral rate of 6.4% for diagnostic audiological assessments. Nine newborns were identified with permanent bilateral hearing impairment. 26 newborns were identified with permanent unilateral hearing impairment. Infants identified with bilateral hearing loss were immediately referred to the Children's Hearing Foundation for hearing aid assessment and fitting. Infants as young as 5 weeks of age were successfully fitted with hearing instruments and enrolled in the family centered early intervention program at the Children's Hearing Foundation. CONCLUSION: The frequency of bilateral congenital hearing loss requiring amplification in this population is shown to be approximately 1 in 752 newborns. This finding is consistent with previous research, which has indicated hearing loss to be the most frequently occurring birth defect. Universal newborn hearing screening using TEOAEs proved to be a cost effective and feasible method of identifying congenital hearing loss in Taiwan. The existence of many successful screening programs worldwide and the availability of fast, objective, reliable and inexpensive hearing screening procedures means that universal newborn hearing screening is becoming the standard of care.

Acoustic Stimulation↗
PubMed2002

Acoustic reflex thresholds at varying click rates in children.

OBJECTIVE: Previous studies conducted with young adults have shown that ipsilateral acoustic reflex thresholds improve with an increase in the click-rate. Older adults also show an improvement in acoustic reflex thresholds with an increase in the click-rate. However, the click-rate induced facilitation (RIF) of acoustic reflex thresholds is significantly reduced in older adults. These results suggest that slower temporal processing in older adults can be documented by measuring the click-RIF of acoustic reflex thresholds. Thus, the click-RIF of acoustic reflex thresholds may prove to be a valuable tool for the assessment of temporal processing disorders in young children. However, click-RIF of ipsilateral acoustic reflex thresholds has not been studied in young normal children. The purpose of this study was to evaluate and document the click-RIF in young normal children to provide normative data for this age-group. METHODS: The study included a total of 30 participants (12 male and 18 female) with normal hearing between the ages of 6 and 10 years. Children with outer, middle ear, or neurological disorders were excluded from the study. Ipsilateral acoustic reflex thresholds were determined in the left and right ears with clicks presented at the rate of 50 and 100 per s. The frequency of the probe tone was 226 Hz and intensity was 85 dB SPL. A statistical analysis using a three factor mixed MANOVA was planned to explore differences between ears, click rates and gender. RESULTS: The MANOVA revealed a significant effect (P=0.0000) for the click-rate. There were no differences between ears or gender and no interactions. The mean click-RIF with an increase in the rate from 50 to 100 per s was 10.50 dB. CONCLUSION: Comparison of these results with previous results obtained from young adults suggests that young normal children show click-RIF that is similar to that apparent in young adults.

Acoustic Stimulation↗
PubMed2002

Characterization of the neuronal changes in the medial prefrontal cortex during jaw movement and eyeblink Pavlovian conditioning in the rabbit.

Medial prefrontal (mPFC) single-unit activity was assessed in defensive (eyeblink, EB) and appetitive (jaw movement, JM) Pavlovian conditioning in the rabbit. Concomitant heart rate (HR) changes were also assessed. In a first experiment robust JM conditioned responses (CRs) were observed to a tone-water (CS+) contingency but not to tone-alone (CS-), indicating discriminative JM conditioning. However, the CS-evoked accelerative HR response was not discriminative. Nevertheless, several single-unit discharge patterns were evoked by both tone-water and tone-alone, many of which were uniquely associated with either the CS+ or CS-. In a second experiment, Three separate stimuli, consisting of tone followed by periorbital shock (tone-shock), water (tone-water), and white noise not followed by shock or water, were presented in the same paradigm. Discrimination of conditioned JM, EB and HR changes were observed, i.e. each of these behavioral responses were uniquely associated with the relevant CS presentation. Conditioned bradycardia was evoked by tone-water during the first training session, which changed to tachycardia with further training. However, conditioned bradycardia was evoked by tone-shock throughout training. Different subpopulations of mPFC cells were activated by the tone-shock and tone-water contingencies, but a small group of cells were activated by both.

Acoustic Stimulation↗