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 for robustness in Monte Carlo studies.

R C Serlin1

  • 1Department of Educational Psychology, University of Wisconsin-Madison 53706, USA. rcserlin@facstaff.wisc.edu

Psychological Methods
|August 11, 2000
PubMed
Summary

Monte Carlo studies help select statistical procedures when assumptions are unmet. This research clarifies Type I error control and sample size determination for robust analysis, improving reliability in simulations.

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

Effect of computer support on younger women with breast cancer.

Journal of general internal medicine·2001
Same author

Patient education in pain control.

Supportive care in cancer : official journal of the Multinational Association of Supportive Care in Cancer·2001
Same author

The consequence of ignoring a nested factor on measures of effect size in analysis of variance.

Psychological methods·2001
Same author

Answering two criticisms of hypothesis testing: a comment.

Psychological reports·2000
Same author

A computer-administered telephone interview to identify mental disorders.

JAMA·1997
Same author

Dimensions of the impact of cancer pain in a four country sample: new information from multidimensional scaling.

Pain·1996

Area of Science:

  • Statistics
  • Computational Science
  • Data Analysis

Background:

  • Monte Carlo studies are crucial for evaluating statistical procedures, especially when assumptions are violated.
  • Previous applications of statistical design principles to Monte Carlo studies have misidentified Type I errors.
  • Accurate error assessment is vital for determining the robustness of analytical methods.

Purpose of the Study:

  • To present a method for controlling the correct Type I error rate in Monte Carlo studies.
  • To describe how to determine the necessary number of iterations for desired statistical power.
  • To derive a confidence interval for the true Type I error rate of a test.

Main Methods:

  • Development of a novel approach for Type I error rate control in simulations.
  • Formulation of a procedure for calculating the required sample size (number of iterations) to achieve specific power.
  • Derivation of a confidence interval for estimating the true Type I error rate.
  • Proposal of a new robustness criterion balancing existing standards.

Main Results:

  • A method is established for accurately controlling the Type I error rate in Monte Carlo simulations.
  • Guidelines are provided for determining the optimal number of iterations for achieving desired statistical power.
  • A confidence interval for the true Type I error rate has been derived, enhancing estimation accuracy.
  • A new, balanced criterion for assessing statistical robustness is introduced.

Conclusions:

  • The presented methods enhance the rigor and reliability of Monte Carlo studies.
  • Researchers can now more accurately assess statistical procedure robustness and determine appropriate simulation parameters.
  • This work provides a framework for more informed decision-making in selecting analytical procedures under non-ideal conditions.

Related Experiment Videos