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

Statistical evaluation of method-comparison data.

G T Wu, S L Twomey, R E Thiers

    Clinical Chemistry
    |March 1, 1975
    PubMed
    Summary

    Statistical analysis in clinical chemistry often assumes normal (Gaussian) distributions. This study highlights the dangers of non-Gaussian data in method comparison studies and proposes a solution.

    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

    Microscopic detection of cadmium oxide particles in lung tissue.

    The Journal of industrial hygiene and toxicology·2010
    Same author

    [Experimental study on antagonism of ginkgolides against effect of platelet-activating factor on lung strip of guinea pigs].

    Zhongguo Zhong xi yi jie he za zhi Zhongguo Zhongxiyi jiehe zazhi = Chinese journal of integrated traditional and Western medicine·1997
    Same author

    [Effect of hemin in treating hemorrhagic anemia and toxicity].

    Zhongguo yao li xue bao = Acta pharmacologica Sinica·1996
    Same author

    Inhibitory effects of 2-[(diethylamino)acetyl]-1,2,3,4-tetrahydro-6, 7-dimethoxyl-1-[1'-(6"-methoxy-2"-naphthalenyl)ethyl]-isoquinoline on isolated guinea pig papillary muscle and heart atrium.

    Zhongguo yao li xue bao = Acta pharmacologica Sinica·1995
    Same author

    Computers should be more widely used in quality control.

    Clinical chemistry·1983
    Same author

    Enzyme-labeled immunosorbent assay for serum ferritin: method evaluation and comparison with two radioassays.

    Clinical chemistry·1979

    Area of Science:

    • Clinical Chemistry
    • Biostatistics
    • Laboratory Medicine

    Background:

    • Method comparison studies are crucial in clinical chemistry for validating new analytical techniques.
    • Traditional statistical methods often assume Gaussian (normal) distributions for the data.
    • This assumption may not hold true for populations of differences or paired values in method comparisons.

    Purpose of the Study:

    • To question the assumption of Gaussian distributions in method comparison studies.
    • To illustrate the risks of misinterpreting results due to non-Gaussian data.
    • To present a systematic statistical approach to avoid these problems.

    Main Methods:

    • Analysis of laboratory data comparing two analytical methods.
    • Examination of published data from method comparison studies (1968-1973).
    • Evaluation of statistical distributions for differences and paired values.

    Main Results:

    • Data indicate that distributions of differences and paired values are often non-Gaussian.
    • Examples demonstrate the potential for incorrect interpretation of results.
    • Failure to account for non-Gaussian distributions poses a significant risk in method validation.

    Conclusions:

    • The assumption of Gaussian distributions in method comparison is often invalid.
    • Non-Gaussian data can lead to erroneous conclusions in clinical chemistry.
    • A systematic statistical approach is necessary for accurate method comparison.

    Related Experiment Videos