Document Detail

The Prep statistic as a measure of confidence in model fitting.
MedLine Citation:
PMID:  18605475     Owner:  NLM     Status:  MEDLINE    
In traditional statistical methodology (e.g., the ANOVA), confidence in the observed results is often assessed by computing thep value or the power of the test. In most cases, adding more participants to a study will improve these measures more than will increasing the amount of data collected from each participant. Thus, traditional statistical methods are biased in favor of experiments with large numbers of participants. This article proposes a method for computing confidence in the results of experiments in which data are collected from a few participants over many trials. In such experiments, it is common to fit a series of mathematical models to the resulting data and to conclude that the best-fitting model is superior. The probability of replicating this result (i.e., Prep) is derived for any two nested models. Simulations and empirical applications of this new statistic confirm its utility in studies in which data are collected from a few participants over many trials.
F Gregory Ashby; Jeffrey B O'Brien
Related Documents :
17639505 - Modelling, prediction and adaptive adjustment of recruitment in multicentre trials.
6987765 - Pitfalls in randomized surgical trials.
18639475 - Carotid endarterectomy: technical practices of surgeons participating in the gala trial.
19279565 - The influence of underlying assumptions on evaluating the relative merits of concentrat...
24932965 - A simple and reliable method to perform biomechanical evaluation of postoperative nerve...
23720545 - Phylogenetic prediction of the maximum per capita rate of population growth.
Publication Detail:
Type:  Journal Article; Research Support, N.I.H., Extramural    
Journal Detail:
Title:  Psychonomic bulletin & review     Volume:  15     ISSN:  1069-9384     ISO Abbreviation:  Psychon Bull Rev     Publication Date:  2008 Feb 
Date Detail:
Created Date:  2008-07-08     Completed Date:  2008-11-05     Revised Date:  -    
Medline Journal Info:
Nlm Unique ID:  9502924     Medline TA:  Psychon Bull Rev     Country:  United States    
Other Details:
Languages:  eng     Pagination:  16-27     Citation Subset:  IM    
Department of Psychology, University of California, Santa Barbara, California 93106, USA.
Export Citation:
APA/MLA Format     Download EndNote     Download BibTex
MeSH Terms
Analysis of Variance
Bayes Theorem
Confidence Intervals*
Data Collection / statistics & numerical data*
Decision Support Techniques
Discrimination Learning
Linear Models
Mathematical Computing
Models, Statistical*
Models, Theoretical*
Pattern Recognition, Visual
Psychomotor Performance
Reproducibility of Results
Grant Support

From MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine

Previous Document:  Three case studies in the Bayesian analysis of cognitive models.
Next Document:  Object detection and basic-level categorization: sometimes you know it is there before you know what...