The CIOS/Comserve Journal Index system provides a way to perform powerful electronic searches of 60 core serials and annuals in the discipline of communication studies and related fields (e.g., journalism, rhetoric, mass communication, social linguistics).

Indexes for a sample of sources are provided to the general public through this interface; however, full network access to the Journals Index is reserved for individuals who are members of the Communication Institute for Online Scholarship. That access is provided through the e-mail interface and the WWW interface to CIOS services. Those accessing this service from a CIOS affiliate institution may perform searches on 30 serials. Institutions that wish to provide institutional access to all the journal indexes are encouraged to consider obtaining the alternative ComIndex product -- a personal computer program that is in use in hundreds of university libraries and academic departments. However, those accessing this service from a CIOS affiliate institution may perform searches on 30 serials.

ComIndex allows unlimited searching of all the journal indexes using an IBM PC or compatible computer. This program operates independently of Comserve and the networks, containing all the journal data and giving access to it through a professional quality search interface. ComIndex provides a full screen menu-driven graphical interface, a built-in help system, and a means of browsing journal indexes and displaying keyword lists, facilities that are not possible to provide within the confines of the Internet Associates of the CIOS qualify for a substantial discount on personal copies of ComIndex.

When searching journal indexes, keep in mind that they represent an international selection and though most titles are in English (some titles are occasionally in French or German), spelling of words may vary within different journals (e.g., "organization" versus "organisation", "labour" versus "labor", "color" versus "colour", etc.). Use the "or" search operator to search simultaneously for variations in spelling (e.g., /labor/ or /labour/) or, when practical, simply search for a common word fragment (e.g., /labo/). Journal searches are insensitive to the case of letters; thus, a search for /LABOR/ would match "Labor" as well as "laBor".

Error Rates. Due to limitations in scanning technology and wide variation in journal typography and layout, each journal data set is constructed manually. The process involvesdata collection, data entry, and several data verification steps. Every effort has been made to assure the accuracy of the journal indexes and to this end, each index is prepared and proofread by disciplinary experts. Even so, some error is inevitable since the original publications from which the indexes were derived are themselves subject to errors of construction (e.g., omitting author's names, incorrect identification of journal issues, etc.), typography, and proofreading. Nevertheless, the total rate of error within the ComIndex data files is believed to be small as the verification process is very thorough.

Special Symbols. The inability to reproduce non-English characters, phonetic symbols, and so on results occasionally in unavoidable minor differences between original sources and the journal data files. In each case, the common ASCII character most similar in appearance to the original character has been substituted or a description of the missing character is described (e.g., the text ``not equals" may appear instead of the not equals sign). Though not common, these problems occur with greater frequency in journals published outside the United States (e.g., European Journal of Communication, Asian Journal of Communication). In the case of accented letters, such as those occurring in French and Spanish, the unaccented letter has been substituted. Words originally appearing in italic or bold face in original sources are enclosed in single or double quotation marks.

Excluded Material. Although book reviews were included in some indexes that were created early in the indexing process, they were omitted in most indexes and, as a matter of policy, will not be included in any indexes in the future. Significant review essays, however, are included. The indexes are restricted to articles that reflect original research or theoretical exposition. Other examples of excluded materials include published bibliographies, software reviews, results of database searches, reports on conferences, summaries of legislative resolutions adopted by scholarly societies, and brief editorial statements introducing special issues of journals.

TIP: In title searches, use wild card searches and search only for short collections of characters -- "rhet*" rather than "rhetoric", "orga*" rather than "organization".

TIP: Attach the conflation symbol (~) to the end of a word stem to obtain all grammatical forms of a word. For example, searching for "argument~" also finds items with "argumentation", "arguments", "argumentativeness", etc.

TIP: In author searches, search for last names only. Journals often list authors by first initial rather than first name.

TIP: Avoid "and" searches (e.g., searching for /media/ and /television/) unless necessary because they are overly restrictive. You can easily omit the references you don't need once you receive the search result.