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

Fitting heteroscedastic regression models to individual pharmacokinetic data using standard statistical software.

D M Giltinan1, D Ruppert

  • 1American Cyanamid Company, Medical Research Division, Pearl River, New York 10965.

Journal of Pharmacokinetics and Biopharmaceutics
|October 1, 1989
PubMed
Summary

This study presents two accessible methods for analyzing pharmacokinetic data with varying response variances. These techniques use generalized least squares or power transformations, adaptable with standard statistical software for robust analysis.

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

Latent factor regression models for grouped outcomes.

Biometrics·2013
Same author

[Emergency medical treatment of palliative care patients at the end of life: regulatory emergency medical care framework--a case control investigation of medical intervention and advance directives].

Der Anaesthesist·2009
Same author

Treatment of palliative care emergencies by prehospital emergency physicians in Germany: an interview based investigation.

Palliative medicine·2009
Same author

[Treatment of oncology patients in the final stadium of disease by prehospital emergency physicians].

Der Anaesthesist·2007
Same author

Simple incorporation of interactions into additive models.

Biometrics·2001
Same author

Kriging with nonparametric variance function estimation.

Biometrics·2001

Area of Science:

  • Pharmacokinetics
  • Statistical Modeling
  • Biostatistics

Background:

  • Heterogeneity of variance in pharmacokinetic data analysis can impact nonlinear regression accuracy.
  • Common methods like weighted least squares or data transformation require careful selection of weights or transforms.
  • Data-driven determination of weighting schemes or transformations is desirable for improved analytical precision.

Purpose of the Study:

  • To introduce two computational methods for addressing variance heterogeneity in pharmacokinetic data.
  • To enable data-driven selection of weighting schemes or transformations within nonlinear regression models.
  • To provide practical implementation guidance using standard statistical software.

Main Methods:

  • Generalized least squares (GLS) approach assuming variance proportional to an unknown power of the mean.

Related Experiment Videos

  • Power transformation applied to both sides of the regression equation.
  • Implementation via standard nonlinear regression routines without specialized algorithms.

    • Main Results:

    • Both proposed methods are computationally straightforward and implementable using common statistical software.
    • The techniques allow data to inform the choice of variance modeling strategy.
    • Sample code for SAS is provided, demonstrating feasibility in other nonlinear least squares packages.

    Conclusions:

    • The presented methods offer practical solutions for handling variance heterogeneity in pharmacokinetic analyses.
    • These approaches enhance the reliability of nonlinear regression by adapting to data characteristics.
    • The methods are suitable for routine use in statistical software packages supporting nonlinear least squares.