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

Testing non-additivity (interaction) in two-way ANOVA tables with no replication.

Aylin Alin1, S Kurt

  • 1Department of Statistics, Dokuz Eylul University, Izmir, Turkey. aylin.alin@deu.edu.tr

Statistical Methods in Medical Research
|February 16, 2006
PubMed
Summary

This study reviews methods for testing interactions between two variables in data analysis. It focuses on techniques for situations with only one observation per cell, where standard methods fail.

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

Prediction of Intraperitoneal Adhesions in Repeated Cesarean Deliveries with Stria Gravidarum Scoring System: A Cross-sectional Study.

Nigerian journal of clinical practice·2024
Same author

Screening for precancerous anal lesions linked to human papillomaviruses: French recommendations for clinical practice.

Techniques in coloproctology·2024
Same author

Can pelvic floor muscle training positions be selected according to the functional status of pelvic floor muscles?

Nigerian journal of clinical practice·2023
Same author

Magnesium yields opposite effects on the nuclear and cytosolic cascades of apoptosis in different rat brain regions.

European review for medical and pharmacological sciences·2022
Same author

Mutation Status and Immunohistochemical Correlation of <i>EGFR</i> Mutations in Gastrointestinal Stromal Tumors.

Balkan journal of medical genetics : BJMG·2021
Same author

Oxidative stress in cows according to calving season: passive calf immunity and its relationship with colostrum quality.

Polish journal of veterinary sciences·2021

Area of Science:

  • Statistics
  • Data Analysis

Background:

  • Testing for interactions between two variables is crucial in statistical analysis.
  • The presence of interaction terms in a model is termed a 'non-additive model'.
  • Standard two-way ANOVA is effective for multiple observations per cell but not for single observations.

Purpose of the Study:

  • To review and present methods for testing non-additivity (interaction) in two-way tables.
  • To address the challenge of interaction testing with only one observation per cell.
  • To provide a comprehensive overview of existing techniques for this specific scenario.

Main Methods:

  • Review of general methodology for two-factor linear models with interaction.
  • Explanation of the two-way ANOVA method for n > 1 observations per cell.

Related Experiment Videos

  • Presentation of specific methods for testing non-additivity with only one observation per cell, including Tukey's test.

    • Main Results:

    • Standard ANOVA is insufficient for single-observation-per-cell scenarios.
    • Specialized methods, like Tukey's one-degree-of-freedom test, are necessary for single-observation data.
    • The paper illustrates these specialized methods with practical examples.

    Conclusions:

    • Effective methods exist for testing interaction with single observations per cell.
    • The choice of method depends on the assumed structure of the interaction.
    • This review provides valuable insights for researchers dealing with limited data points.