ABSTRACT
Many methods have been reported for the management of data collected in the performance of radioimmunoassay (RIA). Data management, sometimes called data processing or data reduction, constitutes an important aspect of the overall RIA methodology and contributes substantially to the success or failure of an assay. Achievement of assay accuracy depends upon avoiding error. It is therefore important not only to avoid error through careful planning, but also to recognize error when it does occur. Imprecision is the result of random error. Random error is usually described quite well by a normal or Gaussian distribution or by a Poisson distribution. Random error is introduced, for example, by the random nature of radioactive decay and by the normal variations expected in the conduct of standard laboratory operations such as pipetting. Bias is the result of a systematic error in the assay method that yields consistently high or consistently low results.
