Hp. Beck-bornholdt et Hh. Dubben, Multiple significance testing and its relevance for interpretation of results. Audit of the Journal "Strahlentherapie und Onkologie", STRAH ONKOL, 176(8), 2000, pp. 344-349
Background: The statistical quality of the contributions to "Strahlentherap
ie und Onkologie" is assessed, aiming for improvement of the journal and co
nsequently its impact factor.
Material and Methods: All 181 articles published during 1998 and 1999 in th
e categories "review", "original contribution". and "short communication" w
ere analyzed concerning the appropriate use of multiple tests.
Result: One hundred and four publications were excluded from analysis, beca
use they did not contain quantitative data or because no or only 1 statisti
cal test was performed. In 77 publications multiple tests were done which w
as adequately considered in only 3 of these papers, corresponding to a frac
tion of 4% (95% CI: 0.8 to 11%).
Conclusion: Authors, peer reviewers, and editors could contribute to improv
e the duality of the journal by setting value on correction for multiple te
sts.