ISSN: 1300-0292
İndekslendiği Dizinler: SCIENCE CITATION INDEX EXPANDED
CINAHL, Index Copernicus,
Chemical Abstracts (CA),
Excerpta Medica / EMBASE
Dil: Türkçe, İngilizce
İçerik: Orijinal Araştırma, Derleme, Editöre Mektup, Olgu Sunumu, Tıp Eğitimi, Tıbbi Kitap İncelemeleri
|
|
|
Bias In Medical Research: Types, Classification, Sources And Control Measures: Review (1)
Dr. Banu ÇAKIRa
aHalk Sağlığı AD, Hacettepe Üniversitesi Tıp Fakültesi, ANKARA
The ultimate goal in medical research is to find the “truth”. Regardless of specific aim(s), researchers strive to calculate point estimates and relevant confidence intervals while studying a sample of a given population. In order to obtain valid and reproducible estimates of the “truth”, researchers should be aware of potential errors and their identifying characteristics and attempt to assess the best ways of avoiding such biases, regardless of the specific objectives and/or sample size of a given study.
Any observed estimate in an epidemiologic study is the result of 3 factors: (1) objective measurement (the ostensible “truth”), (2) random error, and (3) systematic error (bias and/or confounder). Researchers can claim that observed findings reflect the truth only if all such errors are excluded.
These two complementary reviews on bias aim to summarize previous studies on systematic error. This first is intended to acquaint the reader with various sources of potential bias in medical research and to describe their major characteristics under two headings: selection bias and information bias or misclassification.Keywords: Bias, epidemiology, biomedical researchTurkiye Klinikleri J Med Sci 2005, 25:100-110
|
|
|
|
|
