Publication details
Analýza dat v neurologii: XLII. Simpsonův paradox a faktory modifikující účinek v analýze asociačních studií
| Title in English | Data analysis in neurology. XLII. Simpson's paradox and effect modifiers in the analysis of association studies |
|---|---|
| Authors | |
| Year of publication | 2013 |
| Type | Article in Periodical |
| Magazine / Source | Ceska a slovenska neurologie a neurochirurgie |
| MU Faculty or unit | |
| Citation | |
| Field | Neurology, neurosurgery, neurosciences |
| Keywords | Simpson's paradox; data analysis; contradictory outputs |
| Description | Observing Simpson’s paradox indicates severe problems with the study design or effect of a strong confounding factor, which affects results of partial observations and its impact is concealed when merging the individual data sets. Such results must be interpreted with a highest achievable caution, because there is a high risk of bias. |