Friday, March 30, 2012


  • Selection bias, nonrandom assignment to study group, or subjects are allocated in a group without regard their individual characteristics than influenced in the results.
    • The more ill subjects are in one group, the less ill subjects in the other.
  • Recall bias, knowledge of presence of disorder. (see video) 
    • To know to have a disease and don't want to answer questions.
    • To have a children with a congenital disease and recall previous risk factors.
  • Sampling bias, subjects are not representative relative to general population.
  • Late-look bias, information gathered at an inappropriate time.
    • Using a survey to study a fatal disease in a patient who is alive. 
  • Procedure bias, subjects in control and study group does not receive the same treatment.
  • Confounding bias, the effect of one risk factor distorts the effect of the other.
    • Smokes distorts the cholesterol effect (alone) in acute myocardial infarcts.
  • Lead-times bias, early detection of a disease, before it appeared by its natural history. 
    • HIV detection before it become AIDS, treatment changes disease's natural history. 
  • Pygmalion effect, researcher's belief in the efficacy of a treatment changes the outcome of the treatment.
  • Hawthorne effect, group being studied changes its behavior to meet the expectations of the researcher.
    • Patients answer yes to every risk factor to give a positive disease.
  • Detection bias, more information is attempted to obtain from a exposed group than the control group.
  • Allocation bias, subjects are not assigned to study in a non-random fashion.

No comments:

Post a Comment