David P. Roselle, President
104A Hullihen Hall
University of Delaware
Dear President Roselle:
The faculty and staff of the College of Human Resources, Education and Public Policy listed below are greatly concerned about the letter sent to you on February 24, 1999 by the Delaware Association of Scholars. We are dismayed that the Association of Scholars makes general statements about the views and attitudes of University faculty when their own information clearly demonstrates that such generalizations are inappropriate and probably misleading.
The DAS is free to report the responses they received. The fact that the response rate was so low makes the potential for non-response bias almost a certainty. The use of the Roper poli as a bench mark is also inappropriate since the population sampled by Roper was clearly differerent from the one sampled here. There is no basis for saying that, because the results from Roper look like those from the University of Delaware, the Delaware sample is unbiased. In addition, given that there is no way to know if the Delaware sample was by some chance random, there is no legitimate basis for constructing confidence intervals, however broad, or for doing any tests of significance. Further, the pooling of all responses in what was clearly a disproportionate sample (based on the number of responses by college) is simply wrong and the results reported are inaccurate. In short, there is nothing in this work that can be used to generalize to anything beyond the sample drawn.
Even if the study had been executed properly and the results had been generalizable, there are serious issues with the validity of the items measured. The impressions provided by the respondents need to be validated. Are these impressions based on ad hoc experiences? Can one demonstrate that there are measurable declines in any of these areas? If so, can the implementation of policies be clearly linked to the demonstrated declines? If there is such evidence, it is not provided in this study. In the absence of this information, there is no reasonable way that the information provided by the respondents can be linked to any changes in policy. If the research had been properly designed and executed, it might have made some contribution to a serious policy debate. Unfortunately, it cannot.
Sincerely yours,
| Edward Ratledge | Karen Curtis | |
| Timothy Bamekov | Eric Jacobson | Danilo Yanich |
| Deborah Auger | Maria Aristiqueta | Lance Freeman |
| Robert Warren | Donna Bacon | Margaret Wilder |
| Norma Gaines-Hanks | James Flynn | Terry Schooley |
| Kathyrn Denhardt | Leslie Cooksy | |
| Jerome Lewis | Raheemah Jabbar-Bey | Jeffrey Raffel |
| Gwen Brown |