2002
DOI: 10.1191/0962280202sm291ra
|View full text |Cite
|
Sign up to set email alerts
|

Analysis of repeated measures data with clumping at zero

Abstract: Longitudinal or repeated measures data with clumping at zero occur in many applications in biometrics, including health policy research, epidemiology, nutrition, and meteorology. These data exhibit correlation because they are measured on the same subject over time or because subjects may be considered repeated measures within a larger unit such as a family. They present special challenges because of the extreme non-normality of the distributions involved. A model for repeated measures data with clumping at ze… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

1
333
0
1

Year Published

2005
2005
2022
2022

Publication Types

Select...
7
2

Relationship

0
9

Authors

Journals

citations
Cited by 274 publications
(335 citation statements)
references
References 15 publications
1
333
0
1
Order By: Relevance
“…The interaction term, the test of primary interest, tested whether the change across phases differed between the conditions. The viral RNA level, which displayed both an abundance of values below the level of detection and skew in the detected levels, was tested using a two-part model (Tooze et al, 2002) where one of the parts modeled the presence versus absence of detectable viral levels and the second part modeled the mean of the log-transformed values of those levels above detection. This model was estimated twice, once using all assessment points as an outcome and a second time using the two baseline phase assessments (weeks 0 and 4) as covariates.…”
Section: Discussionmentioning
confidence: 99%
“…The interaction term, the test of primary interest, tested whether the change across phases differed between the conditions. The viral RNA level, which displayed both an abundance of values below the level of detection and skew in the detected levels, was tested using a two-part model (Tooze et al, 2002) where one of the parts modeled the presence versus absence of detectable viral levels and the second part modeled the mean of the log-transformed values of those levels above detection. This model was estimated twice, once using all assessment points as an outcome and a second time using the two baseline phase assessments (weeks 0 and 4) as covariates.…”
Section: Discussionmentioning
confidence: 99%
“…46,47 Linkage disequilibrium was expressed as r 2 and D 0 . 48,49 To study the association between ADH1B and ADH1C genotypes and amount of alcohol intake, the correlated mixed distribution model was applied (Mixcorr macro 50 ). This model handles data with clumping at zero and a lognormal distribution for nonzero values, and contains components to model the probability of a nonzero value and the mean of nonzero values, allowing for repeated measurements using random effects.…”
Section: Statistical Analysesmentioning
confidence: 99%
“…This model handles data with clumping at zero and a lognormal distribution for nonzero values, and contains components to model the probability of a nonzero value and the mean of nonzero values, allowing for repeated measurements using random effects. 50 This means that if a variable affects the mean amount by affecting both the probability of occurrence of a nonzero value and also the mean of a nonzero value, these effects can be separated and quantified. Hence, two estimates are produced from this model: the OR for having a nonzero alcohol intake (that is, for not being a non-drinker) and the mean amount of alcohol intake among those with a nonzero intake.…”
Section: Statistical Analysesmentioning
confidence: 99%
“…The first step consists of fitting a two-part statistical model that describes the relationship between usual intake and covariates and estimates the variability of intake both within and between individuals. We adapted a two-part model with correlated person-specific random effects, developed by Tooze et al (3), for this purpose. Similar to the ISUF method, the statistical model represents usual intake as the product of the probability to consume a food on a given day and the usual consumption-day amount.…”
Section: Overview Of the Nci Methodsmentioning
confidence: 99%