Points for Committee Members and Proposal Writers to Consider in the Evaluation of Human Research Proposals

  1. Is this research? Repeats of classic earlier research in classroom settings is usually considered exempt.
  2. How does the combination of the probability of harm to subjects and the magnitude of that harm compare to the combination of the probability of benefit to the subjects, as well as others, and the magnitude of that benefit? This question is what is usually referred to as risk/benefit analysis. If the perception of total benefit does not clearly exceed the perception of total cost, the research cannot be approved. Much of what follows focuses on aspects of that analysis.
  3. Is the study well designed?
    • Is the question clearly stated?
    • Does the design of the study address the question?
    • Will the sampled individuals represent the population from which they are drawn?
    • Are the methods proposed for the analysis of data appropriate?
    • Considering the anticipated magnitude of the differences in response between subject groups, will the proposed sample size be appropriate?
  4. Are the subjects dealt with honestly?
    • Will truly 'informed consent' be obtained? The modern view of 'informed' is that the subjects should be given enough information about the study to get a feel for its possible value and, more importantly, the full picture regarding possible harm to themselves.
    • In situations where informed consent cannot be obtained (e.g.: children and demented patients), will informed permission be obtained from their guardians as well as assent from the subjects themselves?
    • Occasionally, deception of subjects is essential to the design of a study. Subjects may be deceived as to the purpose of the study but not the risks if they are later adequately debriefed.
    • Will subjects be told, in simple language, that they do not have to participate, that they may refuse to answer particular questions, and that they may cease participation at any time without loss of privileges?