Canon argues that the available research, although imperfect, is inconclusive and does not support either proponents or opponents of the exclusionary rule, despite opponents' claims that the evidence indicates the ineffectiveness of the rule. He further notes that his own research, like that of others, used two methods of gathering information on the rule's impact: drawing inferences from recorded data and asking law enforcement officials about their policies, observations, and behavior. He notes that both sets of information produced the same results, despite Schlesinger's view that questionnaire respondents may have lied. In reply, Schlesinger states that proponents of the exclusionary rule bear the burden of proof for proving its deterrent effectiveness for two reasons: The United States Supreme Court's position that deterrence of improper police behavior is the main rationale for the rule and the fact that the rule's benefits must be shown to outweigh its known costs and disadvantages. He further notes that all but one of the seven studies discussed by him and by Canon conclude that the rule does not generally deter. He further states that Canon's own study used a sample which was neither random nor representative and had other methodological problems as well. He concludes that the exclusionary rule results in the release of many dangerous and violent persons and is failing to deter the police from improper behavior. Footnotes are provided.