Zenhausern, 'Listserv Database and the Virtual Personality of a List', Arachnet Electronic Journal on Virtual Culture v2n02 (May 16, 1994) URL = http://hegel.lib.ncsu.edu/stacks/serials/aejvc/aejvc-v2n02-zenhausern-listserv The Arachnet Electronic Journal on Virtual Culture __________________________________________________________________ ISSN 1068-5723 May 16, 1994 Volume 2 Issue 2 ZENHAUSE V2N2 - ========================================================== The Listserv Database and the Virtual Personality of a List Robert Zenhausern St. John's University Jamaica, NY 11439 drz@sjuvm.stjohns.edu Abstract This paper is an introductory tutorial to the statistical and database capabilities of the Listserv which provides basic information about list activity and the Listserv Database program that allows boolian key-word searches of list archives. A second part of the paper demonstrates how this statistical and database information can be analyzed by means of spreadsheet program to determine unique characteristics of a List. The paper closes with the suggestion that this information could be integrated into what could be called the virtual personality of a List. Introduction Computers communication has only gradually emerged as a major force for almost all aspects of life, but computer bulletin boards (BBS) were in existence years before the introduction of the first IBM. Originally they were single user systems, run on public domain or shareware software. These were always open to anyone, free of charge. Gradually multiuser systems based on mainframe computers emerged and some of these were commercial, some academic, and others from diverse areas. Even the earliest BBS had a place for electronic discussions but it was the Listserv software that has the most sophisticated capabilities. Listserv lists are based on the Listserv software that has been under constant development by Eric Thomas since 1986. There have been a large number of Listserv software developed for the Internet, but most do not have the full range of capabilities of Listserv. For example, the Index, Stat, and Review commands provide detailed information on List activity along a continuum of general to specific. Index Command The command: Index altlearn, results in a list of all files that have been added to the List Library and a summary of the List Archives. Many lists do not have a library of added files, but almost all include archives. The archives of any list have the form: altlearn log9204; autism log9301a, etc. That is, the messages posted to the Altlearn List for April, 1992; the messages posted to the Autism List for the first week of January, 1993, etc. These archives are available from Listserv by email to Listserv@site, with the message: send altlearn log9204, and the messages from April, 1992 for autism will emailed. Example 1 shows truncated output of the command: index Altlearn, sent via e-mail to Listserv@SJUVM.STJOHNS.EDU , returns a list of all the files archives that are stored on that list and a listing of the message archives. The name of the archive (eg, altlearn log9204) and the total number of lines are reported. The number of lines per month provides information on overall list activity and can also be used to track the extent of List activity over time to examine seasonal trends. Example 1 Truncated Altlearn Filelist Filename Lines ELC PROPOSAL 382 AUTISM PROPOSAL 326 MOBILITY PROPOSAL 346 NOMORE WRITING 514 TELCOM DISABLED 378 UNPAPER DISABLED 1705 Notebook Archives Lines ALTLEARN LOG9009 27 ALTLEARN LOG9010 4408 ALTLEARN LOG9011 4812 ALTLEARN LOG9012 3031 ALTLEARN LOG9101 12204 ALTLEARN LOG9102 1324 ALTLEARN LOG9103 4254 ALTLEARN LOG9104 5074 Altlearn has 6 files that have been added to the Altlearn Filelist, in addition to the archives from its inception in September, 1990. The Filelist also includes the number of lines for each Log and Example 2 shows the activity (on the basis of lines posted) of Altlearn on a monthly basis for the first 3 years of its existence. The mean number of lines for the first two years remains fairly constant, but the activity in 1992-93 rose dramatically. There is wide variation in the monthly average number of lines over the past 3 years, but, although Altlearn is an academically oriented list, activity does not seem to decrease during the Summer months. Activity is at a low, however, during the Winter Holiday season and Final Examination time. On the other hand, May and June were rather active months, even though they, too, were during Examination time. Example 2 Monthly Average Lines for 3 Years on Altlearn 90/91 91/92 92/93 Mean Month ================================ Oct 4408 1364 6430 4067 Nov 4812 2918 4993 4241 Dec 3031 596 2744 2124 Jan 2404* 374 4433 1602 Feb 1324 4548* 3061 1462 Mar 4254 2814* 1373 1876 April 5074 1893 2756 3241 May 3320 3162 3348 3277 June 1372 696 9403 3824 July 503 6181 5991 4225 Aug 1936 4046 6897 4293 Sept 3294 3936 9047 5426 Mean 2777 2097 5040 * These numbers reflect an average of the 3 year activity. Very large files were sent to the during those months, inflating the monthly total. Stat Command The command: Stat altlearn, sent to Listserv@SJUVM.STJOHNS.EDU returned the total numbers of messages sent to Altlearn and a frequency distribution of the number of messages submitted by each member. The Index command provided information on overall list activity, but the Stat command allows a closer look at the leaders of the List. Example 3 is a frequency distribution of subscribers who sent one or more messages to Altlearn during its existence. Of the 267 members who sent any messages 162 or about 60 per cent sent only one message. It would be interesting to look at this as a measure of lurking and examine differences among lists on this Lurker Index. Example 3 Distribution of Message Frequency Number of messages Frequency 1 162 2-3 49 4-5 22 6-7 10 8-9 8 10+ 16 On the other tail of the distribution, there seems to be a group of 16 individuals who contribute most heavily to list activity and Example 4 provides a detailed frequency distribution of this upper range. These subscribers can be considered the regular contributors to the list and the four individuals who have posted more than 100 messages can be considered leaders. Incidently, two of the four leaders were men and two were women. Example 4 Under 10 251 11-40 9 41-70 2 71-100 1 Over 100 4 Review Command The command, review Altlearn, sent to the Listserv returns of list of the default setting of the List and provides a list subscriber names and userids of current subscribers. A modification of the command, review Altlearn (country, lists the subscribers alphabetically by country and produces a frequency distribution of the number of subscribers from each of the countries Example 5). This command is a Rosetta stone linking userid and the Person. For example, the gender of an individual can frequently be determined by name. More males than females were subscribed to Altlearn with 96 subscribers (53%) male, 72 (40%) female. Gender could not be determined for the remaining 7%. Example 5 Distribution of Subscribers by Country Country Subscribers USA 14 Canada 12 Australia 4 Mexico 3 Turkey 2 Germany 2 Spain 2 Unknown 3 One each from each of the following: Israel, Finland, Iceland, Switzerland, Great Britain, Taiwan, Puerto Rico, Columbia, Hong Kong, Saudi-Arabia, Poland, Chile, Brazil 13 Total from 21 countries 182 The Index, Stat and Review commands provide quantitative information about list activity, but the Listserv Database Program, LDbase, allows more qualitative analysis. The following section provides a short tutorial on searching the Listserv Databases. It will be followed by examples of the statistical analyses that are possible using a spreadsheet program. Searching a Listserv Database The archives of a list can be retrieved very simply, but the particular reference within that archive may not be as easily accessible, especially for large archives. Listserv provides a way to search a List Archive by email or interactively on CMS or VMS systems. Both batch and interactive processing are well documented in the file LDbase Memo which is available from most Listservs. Send this message via e-mail to Listserv: info database and a copy of LDbase Memo will be emailed. Submitting the Search The command syntax and search rule, described below, are embedded in a template (See Example 6) which is mailed to the Listserv at the List node. The examples below are based on the list Altlearn located at Listserv@SJUVM.STJOHNS.EDU Example 6 Job Command Language for Batch Listserv Database Search //Scan JOB Echo=Yes DATABASE SEARCH DD=Data //Data DD * {search rules and commands inserted here} .... .... /* The Search Command A search can be performed on the basis of a string, subject, sender, date, and sound in unlimited boolian permutations. The most basic command is search * in altlearn which would result in the selection of all the messages in the archives. Such a search would have limited use and the search rules can be modified to include only messages sent after a given date, before a given date, or between two dates. Thus: 1) search * in altlearn since 92/01/01 2) search * in altlearn before 92/01/01 3) search * in altlearn from 92/01/01 to 92/06/30 will select all messages sent to the Altlearn List, 1) since January 1, 1992; 2) before that date; and 3) for the first 6 months of 1992. A specific string can be searched, using the command: search facilitated communication in altlearn search facilitated communication in altlearn since 93/01/01 This would provide a list of all messages containing the words "facilitated" and "communication". The search rules can be modified to send messages that contained the string that appeared, for example, since the beginning of 1993. It is possible to select messages on the basis of the sender of the message. The following command would return a list of all the messages sent by John Doe from Syracuse University: search * in altlearn where sender is jdoe@suvm In combination with date, only messages by John Doe for the last half of 1992 will be returned: search * in altlearn where sender is jdoe@suvm from 92/07/01- to 92/12/31 (Note the "-" character is used to indicate a continuation of the command on the next line.) It is also possible to do phonetic searches: search * in altlearn where sender sounds like low The full complement of Boolian keywords (AND, OR, NOT, CONTAINS, etc) are available in the Listserv Database program, and this paper will not go into that detail. The logic, however, is intuitive and, Listdb Memo provides complete documentation replete with examples. The Index Command After the search rules have been formulated the Index command provides a list of the selected messages. The Index includes the number of the idem, and its date, time, number of lines, and subject. Example 7 is the truncated output of the two commands: search * in altlearn index Example 7 Index of Database Search Item # Date Time Recs Subject ------ ---- ---- ---- ------- 000001 92/04/25 07:12 25 A few words 000002 92/04/27 08:56 40 A few words from me too! 000003 92/04/27 10:33 83 Introductions 000004 92/04/27 11:30 21 Re: A few words from me to o! 000005 92/04/28 13:14 36 17 strong and still growin g. This index of messages that meet the search criteria, serves as a guide to which message should be retrieved. The List command can be used to do some rudimentary formatting of the report returned by Listserv. The fields and their widths of the report can be modified according. Of most importance to this paper is the list command can be used to include the sender of the message in the index. The following commands will result in an index, part of which is reproduced in Example 8. select * in altlearn list sender.9 index The command prior to will result in the first 9 characters in the userid of the person who sent the message in addition to the fields usually contained in the index command. Example 8 Truncated List/Index Output Sender Item # Date Time Recs Subject ------ ------ ---- ---- ---- ------- DRZ@SJUVM+000001 92/04/25 07:12 25 A few words RJKOPP@SU+000002 92/04/27 08:56 40 A few words from me too! DOC@VTVM1+000003 92/04/27 10:33 83 Introductions jmwobus@M+000004 92/04/27 11:30 21 Re: A few words from me too! RJKOPP@SU+000005 92/04/28 13:14 36 17 strong and still growing. Print Command The "search" and "index" commands will result in email containing the index of "hits". If the command "print N" is included ("print 26" or "print 27 30-35") the text of the selected messages will be sent. Searching a Listserv Database in Interactive Mode For those actually on CMS/VM or VMS sites it is possible to use the LDbase exec program available from Listserv. To use the program, type LDbase listnode, and use the same commands as in batch mode. One advantage is the ability to ask for specific message numbers immediately. Unless the "print" command is given, however the index is not sent. Qualitative Analysis This section of the paper will be devoted to an examination of the information that is available for a specific period of time, dealing with a specific topic or sent by specific individuals. The File List and Cycles of Activity An index was created of all messages sent to Altlearn@sjuvm.stjohns.edu and the index was parsed into a database, resulting in a database of over 1500 entries (less than 1% were deleted because of intelligibility) with the following fields: Sender, message number, message date, message time, number of lines, and subject. A basic question that applies to all Lists is: Who is sending the messages? Using the Extract Unique command of Lotus 1-2-3 it was possible to identify 91 individuals who sent messages to the List during a 9 month period. The number of messages sent by each of these individuals was determined and the resultant frequency distribution is shown in Example 9. Example 9 Distribution of the Number of Messages Sent by Individuals Messages Frequency Per Cent 0-2 47 48.5% 3-5 20 23.7% 6-8 8 6.2% 9-11 1 1.0% 12-14 1 4.1% 15-17 2 3.1% 18-20 3 2.1% 21-23 4 0.0% 24-26 0 2.1% 27-29 0 0.0% 30+ 11 9.3% Over 70 per cent of the individuals sent less than 6 messages to the list in the 9 month period, 19 who sent between 6 and 23 messages, and 11 individuals sent 30 or more messages. This leads to an initial formulation on List membership, consisting of "one time users", "regulars", and "leaders". Note that this is a more precise measure than the stat command which reflects activity for the complete life of the list. A second basic question that can be asked of any List is how long are the messages? A frequency distribution of the number of lines in each of the 1517 messages is shown in Example 5. Example 10 Distribution of the Number of Lines in Messages Lines Frequency Per Cent 1-10 16 1.1% 11-20 394 28.0% 21-30 399 28.4% 31-40 225 16.0% 41-50 144 10.2% 51-60 59 4.2% 61-70 33 2.3% 71-80 37 2.6% 81-90 24 1.7% 91-100 17 1.2% 100+ 57 4.0% The mode of the distribution is between 11 and 30 lines and the most frequent messages are about a one screen in length. On the other hand, 4 per cent of the messages are over 100 lines in length. These two analyses have examined elementary characteristics of a single list and are merely suggestive of the potential inherent in these techniques. If the same analyses were performed on other lists, what differences would occur? Is Altlearn typical of one kind of lists? What other kinds might exist? Analyses might be done on the time of day or day of month of messages. Are there fewer messages in the Summer when Universities are typically less active and people are on vacation? Previous investigators have examined the nature of electronic conferences and have used terms that might be considered virtual anthropomorphism. For example, the insightful and pioneering work of Hiltz and Turoff (1978) used the term "collective intelligence" and spoke about psychological differences in computer mediated interaction. Sproull and Kiesler (1991) looked at self-disclosure and flaming as measures of varieties of social information and were concerned with electronic group dynamics. Rheingold (1993) used the term "grassroots groupmind" as the title of one of his chapters. The information presented in this article is merely suggestive of the characteristics of any list that could be integrated into the virtual personality of a list. References Hiltz, S.R. & Turoff, M. (1978) The network nation. Addison- Wesley. Reading, MA. Rheingold, H. (1993) The virtual community. Addison-Wesley. Reading, MA. Sproull L. & Kiesler, S. (1991) Connections. MIT Press. Cambridge, MA Thomas, E. (1988) Revised Listserv: Database functions. Revised Listserv: System Reference Library. Release 1.5n. Paris. _____ Articles and Sections of this issue of the _Electronic Journal on Virtual Culture_ may be retrieved via anonymous ftp to byrd.mu.wvnet.edu or via e-mail message addressed to LISTSERV@KENTVM or LISTSERV@KENTVM.KENT.EDU (instructions below) or GOPHER gopher.cic.net Papers may be submitted at anytime by email or send/file to: Ermel Stepp - Editor-in-Chief, _Electronic Journal on Virtual Culture_ M034050@MARSHALL.WVNET.EDU _________________________________ *Copyright Declaration* Copyright of articles published by Electronic Journal on Virtual Culture is held by the author of a given article. If an article is re-published elsewhere it must include a statement that it was originally published by Electronic Journal on Virtual Culture. The EJVC Editors reserve the right to maintain permanent archival copies of all submissions and to provide print copies to appropriate indexing services for for indexing and microforming. _________________________________ _________________________________ _THE ELECTRONIC JOURNAL ON VIRTUAL CULTURE_ ISSN 1068-5327 Ermel Stepp, Marshall University, Editor-in-Chief M034050@Marshall.wvnet.edu Diane (Di) Kovacs, Kent State University, Co-Editor DKOVACS@Kentvm.Kent.edu ____________________________ GOPHER Instructions ____________________________ GOPHER to gopher.cic.net 70 ____________________________ Anonymous FTP Instructions ____________________________ ftp byrd.mu.wvnet.edu login anonymous password: users' electronic address cd /pub/ejvc type EJVC.INDEX.FTP get filename (where filename = exact name of file in INDEX) quit LISTSERV Retrieval Instructions _______________________________ Send e-mail addressed to LISTSERV@KENTVM (Bitnet) or LISTSERV@KENTVM.KENT.EDU Leave the subject line empty. The message must read: GET EJVCV2N2 CONTENTS Use this file to identify particular articles or sections then send e-mail to LISTSERV@KENTVM or LISTSERV@KENTVM.KENT.EDU with the command: GET where is the name of the article or section (e.g., author name) and is the V#N# of that issue of EJVC