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

Transformations of count data for tests of interaction in factorial and split-plot experiments.

Mark E Payton1, Scott J Richter, Kristopher L Giles

  • 1Department of Statistics, Oklahoma State University, Stillwater, OK 74078-1056, USA.

Journal of Economic Entomology
|July 4, 2006
PubMed
Summary

For count data in entomology, standard transformations like square root or logarithm may inflate errors when testing interactions. Untransformed or aligned rank data best preserve statistical accuracy for interaction 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

Right lateral thoracic radiographic metrics help discriminate cranial mediastinal masses and pleural effusion in cats.

Journal of the American Veterinary Medical Association·2025
Same author

A Prospective Cohort Study on the Effects of Repeated Acute Stress on Cortisol Awakening Response and Immune Function in Military Medical Students.

Biomedicines·2024
Same author

Healthcare providers' knowledge and attitudes about overdose prevention sites in Colorado.

Harm reduction journal·2024
Same author

Aphid parasitism in winter wheat fields in a heterogeneous agricultural landscape.

Journal of economic entomology·2024
Same author

Natural Enemies, Mediated by Landscape and Weather Conditions, Shape Response of the Sorghum Agroecosystem of North America to the Invasive Aphid <i>Melanaphis sorghi</i>.

Frontiers in insect science·2024
Same author

Ecological Interactions of Predatory Mites, <i>Cheyletus eruditus</i> (Schrank) (Trombidiformes: Cheyletidae) and <i>Cheyletus malaccensis</i> Oudemans, and Prey, <i>Liposcelis decolor</i> (Pearman) (Psocodea: Liposcelididae), under Different Thermo-Hygrometric Regimes.

Insects·2023

Area of Science:

  • Applied Entomology
  • Biostatistics
  • Statistical Modeling

Background:

  • Count-type variables in entomological studies often require data transformation for analysis of variance.
  • Common transformations include square root, logarithm (log), and rank transformations.
  • The utility of these transformations, particularly for interaction testing, requires careful examination.

Purpose of the Study:

  • To evaluate the effectiveness of data transformations (square root, log, rank, aligned rank) in applied entomological experiments.
  • To assess the impact of these transformations on Type I error rates and test power for interactions.
  • To compare the performance of transformed versus untransformed data in factorial and split-plot experiments.

Main Methods:

  • Reanalysis of field-collected split-plot experiment data.

Related Experiment Videos

  • Simulation study using Poisson distributed errors in factorial and split-plot designs.
  • Examination of Type I error rates and power for interaction tests across raw and transformed data (log, square root, rank, aligned rank).

    • Main Results:

    • Significant interactions were found to be dependent on the type of transformation used.
    • Untransformed and aligned rank transformed data demonstrated superior performance in preserving nominal Type I error rates for interaction testing.
    • Log, square root, and standard rank transformations showed inflated error rates in the presence of main effects.

    Conclusions:

    • For testing interactions in entomological count data, untransformed or aligned rank transformed data are recommended over log, square root, or standard rank transformations.
    • The choice of transformation significantly impacts the reliability of interaction testing results.
    • Further research is needed to evaluate transformations for main and simple effects tests.