CRANV1P1 Documents and Questions chapter Cyril Cleverdon Jack Mills Michael Keen Cranfield An investigation supported by a grant to Aslib by the National Science Foundation. Use, reproduction, or publication, in whole or in part, is permitted for any purpose of the United States Government. - 32 - resulting in a much lower acceptance rate. Had more been submitted, more would probably have been accepted, but absolute perfection could not be achieved unless each author examined every document in the collection himself. The relevance assessments in relation to each question are given in Appendix 3G. The questions The authors apparently found no difficulty in preparing the search questions, and the number received was greater than expected, with each author supplying an average of 3½ questions. Space was provided on the form for four questions, and of the 182 authors who replied, 120 supplied four questions. 40 supplied three ques- tions 18 supplied two questions, and 4 authors only submitted the basic question. Th_e high average, together with the fact that two-thirds of the authors supplied four questions, suggests that some authors could have written more questions, if space had been provided. However, since in practically every case all of the cited papers submitted were assessed as relevant to one of the questions given, so imply- ing that none of the references was included specifically to answer a question which they had not supplied because of lack of space, it is reasonable to assume that four questions represented a near maximum for these authors. The requested distinction between basic and supplementary questions clearly fitted the authors' view of their different problems, and only in one case did an author indicate that two of his questions were equally concerned with his basic problem. The 279 questions finally available for testing comprised 118 basic and 161 supplemen- tary questions. There appeared to be no fundamental difference between the basic and supplementary questions. The set of questions is given in Appendix 3D. The subject areas of the base documents were high speed aerodynamics and aircraft s[OCRerr]ructures. The questions mainly fall into these two areas, but some of the supplementary questions in particular concerned subjects away from the centre of the two subject fields chosen. In aerodynamics, some questions dealt with chemistry of gases, sonic boom, flow in compressors, stability and control, spaceflight re:entry, and heat conduction. The structures questions mostly involved thermal and mechanical deformation and loading, with a few on vibration, effects of noise, and material proper- ties. Some questions involved both subject areas, namely on aeroelasticity and flutter, while there were also some purely mathematical requests. The generality of search questions is largely a matter of degree, but we would say, in the context of an aeronautical research organization, that most of the ques- tions are reasonably precise, asking for a clearly defined part of the subject. There are a few broader questions (e. g. Q.41 "What progress has been made in research on unsteady aerodynamics"): there was one question which was not used in the tests because we considered it might have a hundred relevant documents,, and would prob- ably have retrieved the whole collection. As previously stated, 279 questions were available for searching. Of these, 58 were really two or more questions stated in one, since they had a logical sum relationship. (e. g. Q. 129 "What experimental measurements exist of spanwise and chordwise loadings on swept wings at low subsonic speeds and small incidence") For this reason, most of the tests were made with the remaining 221 questions, although at later stages in the tests ,various subsets of thirty to forty questions were used for various purposes. The composition of these various groups of questions is given in Appendix 3E). Questions varied in length; the search terms ranged from 2 to 15 and the average number of individual search terms in the 221 most used questions was 7.6, median 7.9, and the mode was 7. These figures were obtained at the stage when