Data Selection and Retention - Case A3

Mark Creighton is a technician working in a radiobiology laboratory. His current project entails assaying prostaglandin excretion in rats that have been treated with various radioprotective agents before exposure to gamma radiation. Prostaglandin measurements are taken with an autoanalyzer, a machine that automatically dispenses and mixes reagents used in the assay. In one of the experimental runs, six of the ten assays show a reasonably clear, positive correlation between the dosing of a radioprotective drug and prostaglandin excretion. Four of the assays show no such correlation. Upon hearing the results, Mark's supervisor explains that the assay equipment is highly delicate, and unless conditions are ideal and the reagents very fresh, erroneous results can be obtained. He tells Mark that the results showing no correlation will not be considered in the write-up of the research because it is clear that they are not valid. The supervisor asks Mark to freshen the reagents more frequently and to continue collecting data.


  1. In this case, how should the assay results showing no positive correlation be considered in writing up the research? Should they be ignored? How would you decide?
  2. What criteria should one apply in determining whether experimental results are truly erroneous or whether they reflect an actual phenomenon?
  3. What should Mark do if, after following the supervisor's instructions to freshen the reagents more frequently, he finds that the results continue to be ambiguous with regard to the correlation?
  4. What pattern in obtaining negative and positive results would be supportive of the supervisor's assertion that the equipment and reagents are at fault? What role might the appropriate use of controls play in helping Mark deal with this experimental problem?


School of Graduate Studies

230 Elizabeth Ave, St. John's, NL, CANADA, A1B 3X9

Postal Address: P.O. Box 4200, St. John's, NL, CANADA, A1C 5S7

Tel: (709) 864-8000