Zusammenfassung
Hintergrund: Statistische und biologisch-medizinische Signifikanz sind nicht äquivalent und sollten unterschieden werden. Material und Methode: Relevante statistische Daten der Weltliteratur wurden anlysiert und bewertet. Resultate: Eine halbschematische Abbildung erlaubt einen Überblick über die relevanten statistischen Größen. Schlussfolgerung: Das Signifikanzniveau von experimentellen Daten sollte genau an die klinischen Fragestellungen angepasst werden.
Abstract
Background: Statistical significance and biological/medical significance are not equivalent and should be distinguished from one another. Materials and Methods: Pertinent data gleaned from the world literature have been analyzed and appraised. Results: A semi-schematic display of the available data facilitates an overview of the relevant statistical quantities. Conclusions: The significance level of experimental findings should be gauged in accordance with the clinical requirements.
Schlüsselwörter
Signifikanz - statistische Signifikanz - biologische/medizinische Signifikanz - Mittelwertbildung
Key words
Significance - statistical significance - biological/medical significance - averaging
Literatur
-
1 Eddington A S. The Philosophy of Physical Science. Cambridge; University Press 1939
-
2
Romano P E.
The „Statistical Significance = P ≤ 0.05” Trap. Letter to the Editor.
Ophthalmology.
2002;
109
1949-1950
-
3
Romano P E.
The insignificance of statistical significance: against the tyranny of P ≤ 0.5.
Binoculare Vision.
1985;
1
112-115
-
4
Romano P E.
Editorial: if it’s not „significant”, what is it?.
Binocular Vision.
1987;
2
128, 174
-
5
Romano P E.
Imperial statistics.
Binocular Vision.
1987;
2
128, 174
-
6
Romano P E.
Letter to Editor. Secondary exotropia.
J Pedr Ophthalmol Str.
1994;
31
212-213
-
7
Romano P E.
Editorial: Tschebysheff! Yet another reason you should not use „statistical significance” = a probability ≤. 0.5.
Binocular Vision.
1998;
13
15
-
8
Romano P E.
Why and how we should replace „statistical significance” with „medical/clinical significance”.
Binocular Vision & Strabismus quarterly.
1985 - 1999;
-
9
Romano P E.
The insignificance of a probability value p ≤ 0.05 in the evaluation of medical scientific studies.
J Lab Clin Med.
1988;
111
501-503
-
10
Romano P E.
Abandon Ye the cult of „statistical significance”! - Statistical significance and clinical significance are not the same!.
Binocular Vision & Strabismus Quarterly.
1996;
11
109-110
-
11
Romano P E.
The insignificance of a probability value p ≤ 0.05 in the evaluation of medical scientific studies.
J Laboratory & Clinical Medicine.
1988;
111
501-503
-
12
Roman P E.
Editorial: Thanks and a clarification on statistical significance.
J Laboratory & Clinical Medicine.
1988;
112
498
-
13
Mikulecky M.
Editorial: The scientific truth, the signficance level and the p value of a test.
Laboratory and Clinical Medicine.
1989;
6
759
1 Es gibt allerdings genügend Beispiele, in denen die Statistiker dies nicht gesagt haben.
MD Franz Fankhauser
Lindenhofspital
Bremgartenstrasse 117
3012 Bern
Schweiz
Fax: 0 31-3 12 12 71