Herrmann and Yoder lay some serious charges at the door of psychologists researching the implanted memory' paradigm, and in so doing raise important ethical issues that apply to other researchers studying children's testimony. For this reason alone, their paper is a helpful addition to the debate on researching children's memory (see also Goodman and Tobey, 1994;Westcott, 1994). Speci®cally, however, Herrmann and Yoder argue that the implanted memory paradigm places children at risk' since, ®rst, they may react dierently to the paradigm than adult participants, and second, children may suer psychologically (e.g. with respect to authority ®gures, their self-concept and feelings of helplessness). A further charge is that this paradigm may damage the reputation of psychological research.
Because these issues do apply to the wider research discipline it is important to give them careful consideration, as Herrmann and Yoder have done. In this case, unfortunately, the issues become somewhat confused, because it is dicult to apply some of the particular arguments they present to the young children who have taken part in the implanted memory studies they speci®cally focus on (Ceci et al., 1994a,b). In both of these studies, the children involved were preschoolers aged three to six years, and so concerns about self-concept, for example, are likely to be ameliorated. While this may seem a rather super®cial point it is nonetheless relevant, since the age and developmental status of the child concerned is fundamental to assessments of risk (Thompson 1990(Thompson , 1992)). Having noted this as a particular point of concern, it is necessary to consider each of the charges raised in turn.
CHILDREN MAY REACT DIFFERENTLY TO THE PARADIGM THAN ADULTS
This argument appears to be a truism ± for all sorts of reasons children may react dierently from adults (and from each other), and Herrmann and Yoder discuss several of these. However, as they rightly point out, without follow-up research it is dicult to know the degree to which this is true, and empirically such a comparison would be dicult at the present time given the small numbers of children and adults who have participated in relevant studies (e.g. Loftus, 1993;Pezdek and Roe, 1997).