Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Regression models for mixed discrete and continuous responses with potentially missing values

G M Fitzmaurice1, N M Laird

  • 1Nuffield College, Oxford University, U.K.

Biometrics
|March 1, 1997
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Prenatal protein level impacts homing behavior in Long-Evans rat pups.

Nutritional neuroscience·2015
Same author

Prediction of time-to-attainment of recovery for borderline patients followed prospectively for 16 years.

Acta psychiatrica Scandinavica·2014
Same author

Using linkage information to weight a genome-wide association of bipolar disorder.

American journal of medical genetics. Part B, Neuropsychiatric genetics : the official publication of the International Society of Psychiatric Genetics·2011
Same author

Fitting ACE structural equation models to case-control family data.

Genetic epidemiology·2009
Same author

Obesity and weight gain in relation to depression: findings from the Stirling County Study.

International journal of obesity (2005)·2009
Same author

A new statistical screening approach for finding pharmacokinetics-related genes in genome-wide studies.

The pharmacogenomics journal·2008

A new likelihood method analyzes mixed regression models, handling missing data and improving precision. This multivariate approach offers robust parameter estimates and controls statistical errors in biomedical research.

Area of Science:

  • Biostatistics
  • Statistical Modeling

Background:

  • Analyzing mixed discrete and continuous regression models presents challenges, especially with missing data.
  • Separate analyses of distinct responses can lead to interpretation difficulties, bias, and inflated Type I error rates.

Purpose of the Study:

  • To propose a likelihood-based method for analyzing marginal regression models with mixed discrete and continuous responses.
  • To extend the general location model to accommodate missing responses and leverage multivariate analysis benefits.

Main Methods:

  • Developed a likelihood-based method based on an extension of the general location model.
  • Focused on marginal regression models relating response vector expectations to covariates via link functions.
  • Investigated robustness properties of parameter estimates with and without missing data.

Related Experiment Videos

Main Results:

  • The proposed parameterization yields maximum likelihood estimates robust to association misspecification when no data are missing.
  • Multivariate analysis exploits response vector correlations for more precise parameter estimates.
  • The method addresses challenges of nonresponse and potential bias in separate analyses.

Conclusions:

  • The proposed multivariate approach offers advantages over separate analyses for mixed regression models, particularly with missing data.
  • This method provides robust parameter estimation and better control of statistical errors in biomedical applications.
  • Applied the novel methods to two real-world biomedical datasets, demonstrating practical utility.