Abstract

One hundred eighty trauma-focused dissertations published in the United States were examined to determine the variation in risk language used in the informed consents. Level of risk proposed in the informed consents was poorly related to ratings of risk by graduate coders and virtually unrelated to vulnerability factors such as the age of participants and clinical or nonclinical status. Risk language in the informed consents was markedly elevated over that rated by the coders, with more than one third of the dissertations presenting the risk of participation as severe (i.e., associated with substantial risk of reactivity and lasting discomfort). These statements are not in keeping with the literature and raise ethical concerns in relation to (a) nocebo effects, (b) distortion of correlational findings, and (c) discouragement of participants from disclosing trauma in other settings.