The Variability of Results
Even when the results of a trial are free of bias, the measurements on which they are based are still subject to random variability. Often the variability can be both estimated from its repeatability and reduced by quality control. The measurements may vary between subjects and this variability can be reduced by employing the within subject cross-over trial design. It must also be remembered that a measurement may be very repeatable but either not provide the correct result (low accuracy) or measure something other than that intended (low validity).
KeywordsHeart Attack Blood Pressure Measurement Korotkoff Sound Mercury Column Random Blood Sugar
Unable to display preview. Download preview PDF.