Summary of selected Group 2b study results

 Cited URL connectivity (survey questions I, 1-3)

     The study samples were selected from the CONSER database and the ARL directory, 100 from each. The combined sample size was 197 items, there were three titles that appeared in both the CONSER and ARL samples.  There are still 36 items that have not been returned from surveyors or have not been surveyed.  Of the remaining 161, we were able to view a total of 138 sites.

    Of  the 161 items we surveyed,  60 (37%) of the resources could not be reached by the address given in the CONSER or ARL citations.  Our study didn't require surveyors to give an explanation of why the initial citation did not work, but it appears that at least 28% of these first attempts resulted in not found, unknown host, no DNS entry, access forbidden, and no response messages.  Five samples were titles recently purchased by Kluwer and citations for these still showed the previous publisher's URL's.  Ultimately, we were able to connect with three of these Kluwer titles through further searching.  Miscellaneous reasons for not connecting included problems with the the URL in the CONSER record: not direct enough (e.g. leads to a higher level page, rather than the cited title), an NSDP pre publication record that is still unpublished, cases where the CONSER record didn't contain a URL or citation.  In at least three cases a new page reference was encountered that led surveyors to a new location for the resource.  The actual titles and abbreviated comments can be viewed in Table 1 showing the results of  our attempt to use the cited url (survey question I, 2).

   Ultimately, were unable to connect with 23 of the 161 surveyed titles, about 14%, through the use of the cited URL or any alternative means.   These 23 titles are listed in Table 2.  Our survey did request that a reason be given if we could not connect by any means.  Beside some of the reasons mentioned above, surveyors pointed out that some URL's led to a resource other than the cited title, and the need for a logon or subscription to connect. These are summarized below:
 
 
Unknown host, can't connect, No DNS entry etc. Titles recently purchased from Kluwer URL connected to a different resource Need logon or subscription to continue No Reason given
10 2 5 2 3



 

Title changes (survey questions V, 1, a-d)


    There were a total of 13 confirmed title changes for serial resources (summarized in Table 3).   An additional 4 items represent a change from serial to non serial,  these are listed in Table 4.  You'll see in the comments that some of these have apparently changed from serials to Web resources of a different nature (personal Web pages, commercial services site, etc.) and all have apparently changed title as well. So for our total number of confirmed title changes includes both the titles in tables 3 and 4, yielding 17 title changes, or about 12% of the total titles surveyed (138)

    The 3 items listed in Table 5 show evidence of a title change for the print relative (in an OCLC record for example) but no change yet online. The change doesn't appear online yet because the publisher hasn't digitized back to issues that show a different earlier title.  I didn't count these in the above total of confirmed title changes-- though I am willing to reconsider this.  Some of  these are situations where the title change wasn't so long ago for the print version and surveyors pointed out its likely a title change situation will occur soon online when these earlier titles are digitized and mounted.  This situation wasn't always consistently picked up on the survey forms or in compiling the database used for analysis.  So there may be more cases of this among our sample.  It is certainly something to keep in mind in considering how to handle title changes as many of the commercial publishers have the potential, if not always the motivation to digitize back issues that have or once had a different title. We are also not certain yet how publishers will treat these as far as reformatting them or maintaining separate identities for the changed titles.

     There were 3 titles that appeared to be title changes, but the surveyor was unable to confirm for certain if a change had occurred, summarized in Table 6
A combined list of the titles appearing in the title change tables 3, 4, 5, and 6 appears in Table 7   with links to the resources. (Some readers may need to reload Table 7, it was the first document I made available and it has since changed.)

     In looking at whether publishers reformatted earlier issues with a new title or made both available at one site, I've considered the confirmed serial title change in Table 3 first.   It appears 7 publishers provided issues with respective earlier and new titles at the same site, 3 have reformatted so that the earlier title no longer appears on issues that originally contained it.  The publisher of one title, eWeekly has apparently combined the two, reformatting some issues with a new title, maintaining the old title on other issues.  There are two titles where it is unclear what is happening.

    Surveyor comments from Table 3  include the fact that some services provide clear access to resources that have changed title,  as is the case with the JSTOR title change providing clear separate introductory pages for earlier and later titles at the larger JSTOR service site.  In other cases, title changes have occurred where publishers are inconsistent with title presentations causing us to wonder whether the title has changed or not.  Sometimes the focus of the resource or the publisher changes, causing a title change. Frequently Web design features change that have an impact on title presentation.

    The other category of title changes summarized in Table 4,  are titles that changed into other types of resources- Web services, directories, etc.   Three of these completely changed the title as they changed their focus, one used earlier and later titles on various parts of the reconfigured site.

Best source of title comment summaries (survey question VI, 3)

   Home page as source

Table 8

This  table shows titles that surveyors thought would best be treated by selecting an initial/entry page,  journal home page, or  specific journal home page in a multi-title platform service (such as IDEAL, Project Muse, or Wiley's InterScience site) as source of title proper.  There are 46 titles here.  To briefly summarize some of the comments made by surveyors:

      Nine comments mention that the home page provides a formal title presentation and sometimes fuller title (with subtitle information for example). Along similar lines, three comments note the fact that the home page frequently acts as an obvious, intended central organizing and navigation point for the serial. Also, four comments suggest that use of the home page in some instances keeps the title presentation in sync with the title presented on the print version, when there are several title variants and sources to choose from online.

 Five comments highlight cases where the title was presented inconsistently or not at all  on the issues.  In at least four instances surveyors found non serial resources or other situations where the only choice of title was the home page.  Other comments include the mention of the need for liberal title variant added entries once the title proper is selected.  Title prominence on the home page is mentioned at least once, as is an instance of selecting the home page as the title source to avoid a title change.

  At least fifteen of the titles in the "home page is best source" category are from services providing access to a wide range of serials including IDEAL/Academic, Wiley InterScience and Project Muse.  These services for the most part provide a journal home page specifically for each title.
 

  Issue as source

Table 9

This table shows surveyor comments suggesting that an issue is the appropriate title source for 27 of the sampled resources.   In the related Table 10  seven additional cases are shown where the main entry point for a journal is the initial frames set, often presenting the current issue in frames rather than a separate home page.

   Formality of title presentation on the issue over other sources is mentioned in at least 7 cases.  Eight other comments suggest that an issue was the only source available in certain instances, some mentioning that a home page specific to the serial itself was not available. These include situations where the serial is mounted within a larger organizational Web structure and a case where a site was set up to give access to two different serials (not title change related, two separate serials issued by the publisher.)  These also included cases where the entry point for the serial was the current issue.  Other comments involved situations where the current issue *only* is available at the site and cases where the issue consists of a single file (as an email issue)

  Other comments suggested that in some unusual cases,  the issue and its contents may be more stable than the larger site around it, the specific example being a government document.  In fact, though I'm not suggesting its statistically significant, its interesting to note that 6 Canadian Parliamentary documents, 4 U.S.
government published serials and the government document from Spain ended up on this table.  Some of these sites were obviously undergoing reorganization, some government publishers themselves remarking on the need to constantly update the site.
 

 Current issue and persistent frames

Table 10

All 7 titles on this table have a framed initial entry point, many providing access first and foremost to the current issue within title and navigational  frames.

 Either issue or home page

Table 11

There were twelve comments suggesting either initial home page or issues would be a good source of title.  The theme here is consistent presentation of the title, and situations where the publisher uses the same image file or title logo throughout the publication.
 

 Other sources

Table 12

This table represents 6 cases where surveyors thought a source other than issue or home page was appropriate.
 

 Miscellaneous comments


Table 13

This table shows 16 "other" comments regarding alternate title sources, and comments on other matters.


Summary of title data from survey questions II, III & IV

Initial files or pages found once connected (survey question II)

     This section asked for information about titles found on  the initial pages (home, contents, welcome, etc. pages). 124 of the 138 titles we surveyed had such a page that showed a title, corporate name, welcoming phrase or other information.  About 90% of the responses to the survey prompt to list: "Titles on this page:"  were titles that matched or were variants of the cited title.

     Forty-nine of the initial pages actually were named or had some stated function e.g. "welcome page" (survey question II, 2).  At least 14 of these names were some variation of the phrase "home page" (including "...home" and journal home page), at least 9 others were the words "Journal web site" a specific name Wiley uses to identify journal sites mounted on its InterScience service.  Other results can be viewed in Table 14

     Question II,  b asked surveyors to record the document source title from the initial page.  There were 122 responses for this question.  Of these, 32 were exact title matches with the title listed on the initial page (II, a), about 12 were variations of a name for the page e.g. ...home page, "main page", table of contents, a welcome phrase of some sort.  The remaining were made up of a wide range and number of  title variants of the initial page title, some initial titles presented with subtitle, names of corporate bodies, and other information.
 

Target documents (IV)

      Question IV, 2 asked if the resource contained a page or file which acted as a contents page for several volumes or issues and asked the surveyor to record the title if it appeared here.  105 of the resources contained such a page, 95 of these carried a title for the resource.  Question IV, 3 asked for similar information regarding a page or file that acts as a contents page for a single issue.  There were 77 reported issue contents pages, about 90% carrying a title for the resource.   It must be noted that in both the volume contents page and the issue contents page questions, it was not always clear if the respondent was also referring to the initial page under question II or a separate contents and volume page. So I am not sure how useful these questions are to analyzing title sources.

     Question IV, 6 asked about whether the title of the resource appeared on the article itself.  This question was answered yes in 61 cases.  Nine of the titles recorded here were abbreviations of the cited title or title on the initial page.



 

Print versions (V, 2)


     Seventy-three of the surveyed titles had some sort of print version available (about 53%).     We did not find any cases where a surveyor noticed evidence of the electronic version replacing the paper.   Of 32 cases where it was indicated the online version presented different contents than the paper version, it was frequently noted that the online version provided the current issues more quickly than the print.  Also back issue coverage was mentioned as a difference, in that the online versions only covered a limited number of volumes of the print, normally just the last few years.  It must be noted that  for the questions regarding print versions, as with some other areas there generally was not a systematic or easy means of gathering this information. Evidence that a print version existed came from the fact that a bibliographic record existed, the surveyor had personal knowledge of its existence, mention of the paper version online, etc.



Other possible sections/tables that could be added:

1)      I didn't provide much information from sections IV, 3 & 4 which asked about the titles on the earliest and latest issues.  I'm not sure what to do with this information.

2) List of names publishers use for home pages

3) List of all titles and URLs in the study

What else?



Go to:
Comments: Les Hawkinslhaw@loc.gov (12/24/98)