|Title:||Consistency in choice and form of main entry, 1982 and 1989: A comparison of Library of Congress monograph cataloging with that of the British Library and the national libraries of Australia and Canada|
|Author(s):||Jones, Edgar Albert|
|Director of Research:||Henderson, Kathryn Luther|
|Doctoral Committee Chair(s):||Smith, Linda C.|
|Doctoral Committee Member(s):||Allen, Bryce L.|
|Department / Program:||Library Science|
|Degree Granting Institution:||University of Illinois at Urbana-Champaign|
|Abstract:||Random samples of monograph catalog records were extracted from the 1982 and 1989 Australian, British, and Canadian national bibliographies and paired with the corresponding Library of Congress (LC) catalog records, resulting in six samples of ca. 400 "main entry pairs" each. Using logit models, four hypotheses were tested: (1) That the likelihood main entry pairs being identical in choice and form would have increased between 1982 and 1989; (2) That this likelihood would be highest for the National Library of Canada (NLC)/LC group, lower for the National Library of Australia (NLA)/LC group, and lowest for the British Library (BL)/LC group; (3) That membership in categories of these two variables would not suffice to explain the observed variation in likelihoods: that an interaction effect would also be influencing individual likelihoods; and (4) That the primary cause of the increase postulated in hypothesis 1 would be a decrease in divergent national bibliographic agency (NBA) policies.
Hypothesis 1 was accepted, with the log odds of identical main entries in 1989 (0.6265381) being twice that of 1982 (0.3193534). However, an examination of underlying patterns showed that the hypothesis was not uniformly true for the three agency-pairs at the level of individual bibliographic features (e.g., form of personal name heading). Hypothesis 2 was rejected. The pattern that emerged from the data showed the likelihood of identical main entries to be least in the Australian samples (rather than the British samples), and for the 1982 samples, the likelihood was greatest in the British sample (rather than the Canadian sample). Hypothesis 3 was confirmed in part--the likelihood of identical main entries was greatest in the 1989 Canadian sample--but for the most part the pattern revealed in the data had not been predicted. The Australian samples showed little change between 1982 and 1989, while the likelihood of identical main entries in the 1982 Canadian sample was much lower than expected. Examination of patterns for individual bibliographic features revealed likely explanations for the divergence of the observed patterns from those predicted by the hypothesis. Hypothesis 4 was rejected. NBA policies did not appear to achieve a greater harmony between 1982 and 1989, and the 1988 revision of AACR2 affected the outcome of only 6 out of 1,203 cases in the 1989 samples.
The major factors contributing to the patterns observed in the data appeared to be: the operation of an NLC/LC agreement on the use of one another's personal name headings under certain circumstances (1989 Canadian sample); a greater tendency by NLC to elaborate personal name headings (both Canadian samples); an NLA policy artificially lengthening short titles proper (1989 Australian sample); and the withdrawal of an LC policy providing guidance in the same area (all 1989 samples).
|Rights Information:||Copyright 1994 Jones, Edgar Albert|
|Date Available in IDEALS:||2011-05-07|
|Identifier in Online Catalog:||AAI9512418|
Files in this item
|(no description provided)|