Saying it Wrong with Figures: A Comment on Zeisel

Because of the respect and admiration we have for Hans Zeisel's long and distinguished career, it is especially troubling to have to respond to a paper which can only subtract from the sum of his accomplishments. But there are so many errors and misunderstandings in his article that we are forced to point them out lest readers of his critique be led astray. To begin with, there are several misstatements in the article, some more important than others. These follow. (1) Zeisel did not ask us to add a statement in the acknowledgments of Money, Work and Crime recording his disagreement. Instead, he asked that his name be withdrawn from the list of members of our advisory committee in the volume's acknowledgments. The statement was written by us out of a desire to avoid rewriting history, because he had served long and faithfully on that committee and because we wanted to accord appropriate respect to his disagreement. (2) Instead of four payment groups, as Zeisel appears to believe, there were only three in the experiment (a point which is repeated many times throughout the monograph). (3) At many points throughout the monograph we emphasize that the payments as administered had no effects on recidivism. The quotation presented by Zeisel as our statement of the findings is followed in the original text by qualifying statements that clearly point out the zero-order null effects of the payments as administered. Contrary to Zeisel's claim, neither our monograph (Rossi, Berk, and Lenihan 1980) nor our article in the American Sociological Review (Berk, Lenihan, and Rossi 1980) contained attempts to obscure the finding that the treatments failed to achieve their desired effects. (4) The experiment did not test payments, as Zeisel suggests, but only payments as administered.