Charles H. Davis

The Modern Language Association: electronic and paper surveys of computer-based tool use

Members of the Modern Language Association of America (MLA) were surveyed about their use of computer-based tools. A questionnaire was sent to 1,000 randomly-selected members in the U.S., with 500 sent via paper mail and 500 through electronic mail. Word processing, electronic mail, on-line catalogs, and the MLA International Bibliography were used heavily. Responses by the two subgroups differed significantly in several respects. Electronic full texts received substantially less use by both groups, especially those responding to the print survey. Major changes in research habits included greater reliance on word processing and more work outside of libraries. Problems reported focused on access to computer-based resources, learning to use them, the need for instruction, and inconsistent interfaces. Finally, evidence strongly suggests that reliance solely on electronic surveys may produce misleading results.

Entropy and information: A multidisciplinary overview

The concept of entropy, from the second law of thermodynamics, has been used by numerous writers on information theory. Basic relationships between entropy, order, information, and meaning have been observed by writers in disciplines as diverse as biology, economics, information science, the arts, and religion. This article, while not attempting comprehensive treatment, cites representative extensions of the concept, including its use as metaphor.

Collection Overlap as a Function of Library Size: A Comparison of American and Canadian Public Libraries

Random samples of monographic materials from American and Canadian public libraries were matched with the collections of similar libraries of different sizes. Materials with a publication date earlier than 1970 were excluded in order to reduce the sample size, to minimize the effects of comparing older and newer libraries, and to increase the relevance of the findings for resource sharing through contemporary on-line computer-based networks. A strong positive correlation appears to exist between percentage find rate and target library size, with regression analysis suggesting linearity over a wide range of collection sizes.

Computer Programming for Librarians.

Over a five-year period in a graduate library school, 334 students from a variety of backgrounds completed a data processing course involving PL/I programming for library applications. Men and math/science majors enrolled in numbers significantly higher than would be expected considering the composition of a typical incoming class in library school. There was no significant difference in the performance of men and women, but math/science majors performed significantly better than others. There is also evidence that education majors may experience difficulty with a course of this type. An overview of the more popular programming languages is presented along with the reasons for choosing PL/I.

A model for estimating the occurrence of same-frequency words and the boundary between high- and low-frequency words in texts

A simpler model is proposed for estimating the frequency of any same-frequency words and identifying the boundary point between high-frequency words and low-frequency words in a text. The model, based on a maximum ranking method, assigns ranks to the words and estimates word frequency by the formula: Int[(-1 + (1 + 4D/I n+1 ) 1/2 )/2] > n * ≥ Int[(-1 + (1 + 4D/I n ) 1/2 )/2]. The boundary value between high-frequency and low-frequency words is obtained by taking the square root of the number of different words in the text: n * = (D) 1/2 . This straighfforward model was used successfully with both English and Chinese texts, demonstrating that the frequency of words and the number of same-frequency words are dependent only on the vocabulary of a text (the number of different words) but not on its length. Like Zipfs Law, the model may be universally applicable.

Systematic weighting and ranking: cutting the Gordian knot

A powers-of-two algorithm is described that automatically creates discrete, well-defined, and unique result sets, displaying them in decreasing order of likely relevance. All computations are transparent, and a simple query form allows the searcher to focus on the choice of terms and their sequence-an implicit indicator of their relative importance. The program can be used with traditional databases or with search engines designed for the Word Wide Web. It also can be used with an intelligent agent to search the Web with a pushdown store, returning only those items that best reflect the searchers stated interests.

Evidence of OCLC's Potential for Special Libraries and Technical Information Centers.

Searches of the OCLC data base revealed that 65% of a Canadian special librarys recent acquisitions had already been cataloged. This “find rate” compares favorably with statistics available from similar libraries in the United States, and suggests that technical information centers and special libraries throughout North America could benefit considerably from OCLC participation.

Programming Languages Taught in Library Schools, 1980 versus 1986

During a period when microcomputers assumed a major role in contemporary society, there appears to have been a small but statistically insignificant percentage increase among schools that either required computer programming or made it available. The overall pattern for the types of languages taught did not change, but there was significantly more use made of the languages BASIC and Pascal, and less of PL/I and COBOL. There was little reported use of string-manipulation languages such as SNOBOL in the more recent survey, and dBASE and Prolog were mentioned for the first time.

Comparison of recent acquisitions and OCLC find rates for three canadian special libraries

Acquisitions of three Canadian addictions libraries from April 1977 to February 1978 were searched on the OCLC database. Of the 601 titles, 422 were found, giving an average find rate of 70.2%. Find rates for the individual libraries ranged from 67 to 74%. These find rates compare favorably with those reported for U.S. special libraries. Three distinct patterns are revealed for the subject breakdowns of each librarys acquisition. The two smaller libraries-in British Columbia and Alberta-emphasize either alcohol or drugs with nearly equal emphasis on secondary social science, health, and science materials. The largest library-in Ontario-shows a balance among alcohol, drugs, and secondary materials. When each library was compared with the others, the chi-square test for independence showed a difference at the 0.01 significance level. Seventy-six percent of the items searched were U.S. materials; 19.5% of all items were Canadian titles and constituted 80% of all non-U.S. materials. Nearly 10% of all items found were Canadian titles. The find rate for Canadian materials in OCLC is substantially lower than that for others. However, with nearly four-fifths of their acquisitions from non-Canadian sources, these three Canadian libraries could clearly benefit from using OCLC for verification and cataloging.

Relative Percentages of Fiction and Non-Fiction in Selected Canadian and American Public Libraries

A preliminary examination of American and Canadian public library acquisitions published during the 1970s suggests a model in which the percentage of non-fiction increases with collection size. Random samples taken from these particular Canadian libraries have a significantly higher percentage of fiction than the samples taken from their American counterparts; however, the observed difference can be explained by the aforementioned model, in which the variation is a function of collection size rather than nationality.