26.16 Bias

  • Bias – a flaw in impartiality that alters the manner in which measurement, analysis, or assessment is recorded, conducted, or interpreted.
  • Randomization – controls the bias that would otherwise be introduced by confounding variables
    • other types of bias exist
  • Blinding – controls bias that arises when experimenters have information about study subjects that may compromise their impartiality
  • Strict adherence to the study protocol minimizes bias
  • Selection bias is introduced by a nonrandom selection of a sample from a given population
    • Example: only athletes used in the study. They recover quickly.
  • Observational or informational bias occurs when a measurement/outcome is affected by the characteristics of the study group itself
    • Example: survey mailed to participants; some are homeless.