A Response to Caviedes

A Response to Caviedes

W

e too were concerned about the accu-

racy of the index and abstract data in our count of FADS (foreign-area dissertations) in geography. As noted, we used a control of eight programs, sending them the results of our FAD count for verification (Turner and Varlyguin 1995, 310). Of these, six responded, constituting 243 entries of the total 474 FADS in our count, indicating a total number of three errors. Caviedes’s correction, therefore, raises important questions. Was our low level of error a product of the control sample? Or is the Florida correction an anomaly for the reasons noted by Caviedes? We have no reason

to believe that the control sample was biased and conclude that the degree of error (36%) between index-abstract and actual counts in this case is most likely an anomaly. We believe an actual count by all geography programs would produce little change in our overall assessment of trends in FADS for geography at large, although our Table 3 would require several adjustments.

B. L. Turner I1 Dmitry Varlyguin Clark University

Professional Geographer, 48(2) 1996, page 223 0 Copyright 1996 by Association of American Geographcra. Published t)y Blackwell Puldishers, 238 iMam Street, Camlxidge, MA 02142, and 108 Cowlcy Road, Oxford, OX4 IJF, UK.