Indirect evidence of reporting biases was found in a survey of medical research studies.