Zane L. Berge, Ph.D.

Director, Training Systems

University of Maryland Baltimore County

5401 Wilkens Avenue

Baltimore, MD 21228

410-455-2306

FAX 410-455-3986

email: berge@umbc2.umbc.edu

Mauri P. Collins, MA

The Pennsylvania State University

email: mmc7@psu.edu

6/23/95

Interpersonal Computing and Technology: An Electronic Journal for the 21st Century (IPCT Journal) began its third year of publication in January 1995. The Table of Contents file published in the April 1994 issue contained a readership survey designed with several purposes in mind. We wanted to gather further information on several issues of general concern to the scholarly community regarding electronic publishing, authoring, and reading (Berge, 1994). We sought to add to our understanding of authors' and readers' concerns with the credibility, access, and permanence of electronic journals (ejournals) within a changing technological environment and will discuss those issues below.

We were also interested in determining the level of the acceptance of ejournals by academic promotion and tenure committees. In order to maintain a readership, a journal must publish articles of value and relevance. Unless authors are duly rewarded for their writing efforts with credit towards promotion and tenure, or recognition from their audience from a credible, permanent, and accessible platform, they are unlikely to submit their work to electronic journals and will seek print outlets instead.

This readership survey sought to explore these issues by asking our readers their opinions,

with emphasis on their acceptance of ejournals in general and IPCT-J in particular. We asked those in higher education about the role of ejournals in the promotional and tenure process at their institutions. Other questions sought their views on credibility, access and permanence. We also asked a number of questions to give us an indication of the demographic characteristics of the IPCT Journal readership.

As we have used electronic surveys to gather preliminary information (Collins & Berge, 1994b, 1994c, 1994d), we realize that their use is as much in its infancy as are ejournals themselves (Kielser and Sproull, 1986). We felt an electronic survey to be an appropriate method to solicit the opinions of an ejournal’s readership, and realize that our findings cannot be generalized beyond that population. We attempted a census, by sending our survey to those venues that ordinarily receive the journal’s table of contents. We started off with no demographic information, beyond our own observations of the kinds of people who contribute to the ejournal’s companion electronic discussion group Interpersonal Computing and Technology List (IPCT-L). The ejournal subscription list was slowly growing, but we knew that our General Editor felt there was a problem soliciting a sufficient number of quality manuscripts. We realize this sketch of our readership may be skewed or biased, but it is a beginning, and provided us, as publisher and managing editor, with valuable information.

The demographics of the electronic community as a whole is changing from an academic and scientific preserve to a public community as more and more people access the internet via the public providers (e.g., Compuserv, America OnLine, etc.) This may, or may not, change the demographics of ejournal readership, although it is hard to imagine a great public demand for the kind of publications now largely read by scholars and students with specific interests in discipline and domain.

Future surveys will be targeted at a random sample of subscribers and be more rigorous than this exploratory effort in both sampling techniques and analytical methods. We have already begun to work with other ejournal publishers so we can eventually build up a more complete picture. Our efforts in this paper are to develop a general overview first, followed by more targeted inquiry with a better chance of knowing, or calculating, such things as sample size, response rates, and sampling bias.

We started our explorations with an extensive review of the literature. While we found a sizable amount of literature regarding ejournals (e.g., Amiran and Unsworth, 1991; Astle, 1989, Harnad, 1991; McMilan, 1991; Okerson, 1991; Wilson, 1991), we found only two brief summaries of prior ejournal readership surveys. One was a technical report from Syracuse University, "The Electronic Journal: Promises and Predicaments" (Technical Report Number 3) (Ehringhaus, 1990), which included results from an exploratory survey of New Horizons in Adult Education, and a survey of Ejournal readers (Jennings, 1993) conducted in December 1992 and reported in the September 1993 issue (v3n3).

We also looked for readership surveys of paper journals and found several dozen (e.g., Ahmed and Paquet, 1994; Posner, 1987). The purposes for conducting these readership surveys were many and included gathering data about bolstering advertising sales, pre-positioning the publication for increased readability, shaping the future of related association programs and services, reader profiles brand awareness, competitive publication analysis, and many other reasons (Jacobson, 1988; Stratton, 1992). These readership surveys were, however, mostly business, commercial, or professional journals, with few appearing to be peer reviewed.

The mission of the Center for Teaching and Technology (CTT) at the Academic Computer Center, Georgetown University is to promote the integration of computers and other instructional technology in the higher education classroom. Included within that framework is the vision of tapping expertise available from scholars worldwide, and to provide a venue in which they can discuss their common interests, learn from one another, share their expertise and pool their experience toward the solution of common problems.

During discussions with stakeholders at Georgetown University, the publication of an scholarly, peer-reviewed ejournal was suggested as the most immediate and practical way of promoting the mission of the CTT. To that end, a scholarly discussion group, the Interpersonal Computing and Technology Discussion List, (IPCT-L@guvm.georgetown.edu), was created to help to identify an editorial board, potential contributors and subscribers. The scope of topics in the IPCT Journal follows largely those articulated by members of IPCT-L: the use of electronic networks in the classroom; electronic publishing; the use of electronic networks and information exchange; library applications of electronic communication; professional relationships carried on via electronic communication; and perspectives on the use of electronic communication in higher education, business, industry and government. Technological articles are considered to the extent that they are intelligible to the general reader and are not specific to any particular hardware configuration. A Table of Contents containing article abstracts and retrieval information is disseminated four times a year to the IPCT-J and IPCT-L lists, and other lists, selected with a goal of building the subscription list.

In the "Letter from the Publisher" in IPCT Journal 2(1), January 1994, Berge addressed the concerns of credibility, access and permanence of ejournals. Credibility comes largely from the readers' perception of the effectiveness of the peer review process and was looked at in the survey (See Appendix) as indicated by the readers' citation of articles, their willingness to pass articles on to colleagues, their printing out of articles for future reference, and their rating of the quality of IPCT Journal articles.

The credibility of ejournals is a matter of concern to both the authors and members of promotion and tenure committees. In an effort to gauge the acceptance of ejournal publications, respondents in higher education were asked how they would characterize the behavior of administrators and member of committees who decide on retention, promotion and tenure at their institutions regarding electronic publication of refereed journals such as IPCT Journal compared to refereed print journals. They were also asked if they thought electronically published articles should be weighted equally with articles in print journals by retention, promotion and tenure committees.

Technological access issues were addressed in several questions dealing with formats (e.g., ASCII, PostScript, word processor files) and storage media (e.g., gopher, CD-ROM, FTP sites, WWW), and sources of the Table of Contents. Overall awareness to IPCT Journal is greatly helped by its indexing in ERIC and other databases. Acceptance of ejournals was examined in a question asking the subscribers how many other ejournals besides IPCT Journal respondents held subscriptions.

We believe that this type of information could help us and perhaps other ejournal publishers in several ways: to determine the demographic characteristics of our readership, as information if we approach professional organizations for funding support for IPCT Journal, to target marketing efforts for new contributors and subscribers, to make decisions about where to abstract or index the Journal, and to begin to research and understand better this emerging channel for scholarship.

The general characteristics of the IPCT Journal readership were investigated in terms of computer and email expertise, age, gender, ejournal readership, criteria for choice of articles to be retrieved, country of residence, preferences for formats and storage media, and retrieval formats, work organization and their field and specialty. Other questions on the IPCT Journal readership survey address issues general to this genre: retrieval methods, general acceptance of ejournals, source of the Table of Contents received, number of articles likely to be retrieved, criteria for the choices in selecting articles, and whether subscribers had any expectations of the journal before receiving their first issue and how those expectations have been met.

We started this project with only the vaguest idea of who the readers of the IPCT-J actually were. We assumed they would be like, in many respects, the person who post to the IPCT discussion list, as we had originally planned that subscribers to the IPCT-L list would form the basis of the IPCT-J readership. Our assumptions were based on the observations and generalizations that we, as moderators, have made about the IPCT-L subscribers, from three years of reading their posts and their appended signatures.

This survey is a first census of our readership. We decided to send the survey to both IPCT lists and to some of those lists to which the Table of Contents had been posted in the past. The survey was included in the Table of Contents for the April 1994 issue of IPCT Journal. This was posted to the IPCT Journal subscription list (N=1604 in 45 countries at the time of the census) and the related IPCT Discussion List (N=1118 in 48 countries at the time of the census). There is considerable overlap in membership between the two lists so some persons would have received the Table of Contents twice. Additional respondents came from cross-postings to approximately 15 other discussion lists.

Informal interviews were conducted by the Publisher and Managing Editor of IPCT Journal with selected members of the editorial board and reviewers of the IPCT Journal asking for suggestions for survey questions. After the survey was drafted, this pilot document was sent to the entire Editorial Board (approximately 30 people), requesting that they complete it and suggest any modifications. The final readership survey that was distributed drew on the experience of the authors and reflected the responses and suggestions from that pilot group.

The Table of Contents, including the Readership Survey, for the Spring 1994 issue was distributed on April 28th, 1994 to the two lists: IPCT-L and IPCT-J. It was also posted or crossposted to about 15 other lists over the next week (e.g., DEOS-L@PSUVM.PSU.EDU; PACS-L@UHUPVM1). Certainly there is some overlap in subscription to these lists, especially between IPCT-J and IPCT-L, but the exact extent is not known. The survey was redistributed to only the IPCT-J and IPCT-L lists on 5/9/94. Since subscribers to these lists are free to come or go at will, there was some change in the list between the two dates the survey was posted. For example, IPCT-J on 5/9/94 had 1585 subscribers in 46 countries and IPCT-L had 1104 subscribers in 49 countries. While the changes between 4/28/94 and 5/9/94 may be greater than simply subtracting the gross figures for the two days because of the normal comings and goings of subscribers between these dates, we did not feel this to be a significant factor in this study. Because of the changes in subscription to the lists involved, the overlap in individual subscribers among these lists, and the wide distribution of the survey among non-IPCT Journal subscribers, it is felt that any report of response rate would be inaccurate. However, there were 258 respondents who reported receiving their copy of the survey from the IPCT-J list, and that number is 16% of the total subscribers to that list.

It is interesting to note the timing of survey responses compared to non-electronic mail surveys. This is evident in the sometimes rapid turnaround in response, and the relative short duration from the beginning of the data collection period until the end. The Table of Contents, including the survey, was first posted at 8 PM (Eastern Standard Time) from Georgetown University on 4/28/94. By 8:20 PM, twenty minutes later, the first completed survey was returned. Within 24 hours after the initial publication, 98 surveys had been returned. By the end of seven days, 178 surveys had been received. After 10 days, 191 surveys had been returned--only 13 in the combined three days 8, 9 and 10 after publication. On the morning of the 11th day, the survey only (i.e., not the entire Table of Contents) was redistributed to IPCT-J and IPCT-L lists. Twenty-three minutes later the first return came in, from France. During the next 12 hours (i.e., by day 11 from the first date of publication), 93 additional surveys were returned for a total of 284. Thirty-eight days after the first publication of the survey, on June 5, 1994, the last survey was received (number 397). From these, 394 usable surveys were coded and analyzed.

We have used electronic surveys several times in the past (Collins & Berge 1994b, 1994c, 1994d) and determined some of their advantages and limitations. The questions must be just as carefully developed and designed as paper-based surveys, with particular attention paid to how they will look on a screen that is 80 characters wide by 22 lines long, and the ease with which they can be completed. Advantages of esurvey use includes generally rapid response time, which can be in terms of minutes, hours and days, rather than days and weeks. They are also inexpensive to distribute, especially from an academic setting where the cost of email is paid by the institution. Kiesler and Sproull (1986) sum up their research stating:

Rather than sampling, it is often as easy to census the entire subscription list of a discussion or distribution list by sending a single message to the mailing list, which will, in turn, delivers a copy to each electronic mailbox address on the subscription list. Reminders can posted just as easily. Because each survey is returned bearing the email address of the sender, it is a simple matter to courteously send a ‘thank you’ for each response to acknowledge its receipt, and to follow up with further questions to perhaps clarify responses, or elicit further information.

Esurveys have some distinct limitations. Filling a survey out on screen may involve technical expertise that some potential respondents do not have, so the results may always be biased to those who are more experienced or more confident in the use of online text editors and mail systems. It is almost impossible to preserve anonymity as most email programs tag each outgoing address with the address of the sender, and security systems discourage circumventing this by users. We have received responses that have been printed out, all identifying addresses removed, and then postal mailed back to us in envelopes with no return address. We assume that these persons either cannot complete the survey onscreen, or wish to do so while preserving their anonymity.

Because of the ease of distribution, there is a tendency to broadcast esurveys. This may be all right for initial investigations where the researcher may start with little or no information about a given population, as was our case with this study. However, it is then impossible to determine response rates, sampling bias and error, etc. Sometimes the subscription list may include addresses such as "netnews@site.domain" This indicates that the subscriber is actually a mailserver for a site’s internal news dissemination group, the readership of which is impossible to determine. For instance: a message received by netnews@psuvm.psu.edu is posted to the Pennsylvania State University’s internal netnews distribution system, which has a potential readership of more than fifty thousand students and faculty. When sampling, we would suggest such addresses be removed first.

Generally, Shaw and Davis (1995) concluded:

Data was coded and frequencies calculated using SPSS, Version 4, for the 394 usable survey responses. The survey is included at the end of this paper. We have inserted the valid percentages and the (N=nn) for each item that was quantified. At the end of this article you will find our survey instrument with the frequency percentages filled in. We have done this to show the survey instrument and to make the frequencies available to our readers, without cluttering this text with tables.

Readers who responded came from approximately two dozen countries, with approximately 72% of the responses from North America (see Question 6). While not all countries were represented, the percentage of respondents from outside North America corresponds with the percentage of both IPCT Journal and IPCT-L international subscribers.

IPCT Journal readers considered themselves expert in both computers (see Q1) and email use (see Q2). When asked to judge their level of computer expertise, 79.1% reported themselves as "advanced intermediate" or "expert." Similarly, in using email, 76.1% reported themselves in these categories . Only 1.5% and 2.5% considered themselves to be novices in these two areas respectively.

The largest reported age (see Q3) category was 40-49, (44.4%). Slightly over 66% of the IPCT Journal readers reported that they were 40 years of age or older . At first, this seemed surprising, given the average age of computer users is thought to be much younger. However, on second thought, there are not too many teenagers or young adults who might find scholarly journals interesting. Nearly 71% of the respondents were male (see Q4). Since we didn't find much published data on the characteristics of scholarly journal readers, we can only hypothesized that the demographics of typical print scholarly journal readers is more likely to predict IPCT Journal readership demographics than computer users' demographics generally. Slightly over 85% of the respondents reported subscribing to other ejournals (see Q5) besides IPCT Journal, with 15.1% saying they subscribe to an additional five or more ejournals.

With regard to the type of organization our readers belong to (see Q16), 70.2% stated that they are in higher education. An additional 6.7% of the readers responding to this question are in K-12, with only 7.7% in business, and 5.1% in government. We asked the survey recipients in an open-ended question, to list their field and specialty (see Q16). While we made some arbitrary decisions when tallying these, by far the largest group of the 375 persons answering this question listed their field as education (29%). The next three categories were: librarianship (20%), computer systems, science, or user support (15%), and information technology, information science, or instructional systems development (12%). Eighteen persons listed their major field as communication or media theory, and 3% listed business or information management. There were approximately 20 other major areas listed with under 10 respondents each, ranging throughout the humanities, social sciences and natural sciences, (e.g., chemistry (.02%), English (1.3%), health sciences, medicine, or nursing (1.6%), law (1%), political science (2%), psychology (1%), sociology (.8%)).

IPCT Journal readers were asked (in an open-ended question) what they expected the journal to be when they subscribed (see Q14), and how well the journal meets their expectations (see Q15). With regard to readers' expectations before seeing their first issue of IPCT Journal, most responses fell into one of the following categories:

The responses to an open-ended question concerning how well IPCT Journal met their expectations fell largely into the following categories:

While only two dozen or so readers mentioned negative things, you are cautioned regarding the biases inherent in this type of readership survey, (see "Limitations" section below):

It is difficult to determine the number of times a particular journal article is read. The estimate that the average scholarly article is read fewer than 6 times may be part of the folklore, but there doesn't seem to be any solid studies that support this figure (Odlyzko, 1994).

Garvey and Griffith (1971) tried to estimate the number of times articles in core psychology journals were read within 2 months after publication:

Separately, Merton (1968) states that, "only about half of 1 percent of the articles published in journals of chemistry are read by any one chemist" (p. 59).

One assumption made in designing the IPCT Journal is that scholars read articles, not journals. This would be due mainly to the range of topics in a journal being broader than a particular individual's interests. To gather data on this, we asked our readers how many of the six articles featured in this issue they would obtain in full text (see Q8). Only 10.1% stated they would retrieve all of them. Twice that number (21.0%) stated they would not be retrieving any of the articles. The majority of respondents (51.0%) stated they would obtain 1 or 2 articles in that particular issue. A similar pattern was seen when readers responded to how many articles have they generally requested from prior issues of IPCT Journal (see Q8), with 40.3% stating they generally obtain 1 or 2 articles.

Charles Bailey (1994), University of Houston Libraries and founder/publisher of PACS Review writes:

Given the above, we suspect that the average refereed articles in ejournals may have as wide or wider readership as the average refereed printed journal article. As Bailey (above) notes the number of times an article is retrieved is not a completely accurate estimate of the number of times it is read as 53.2% of our respondents said they had sent articles from IPCT-L to colleagues (see Q19). It is also difficult to take into account differences in subscriber numbers, subscription fees, and other factors that may influence this type of comparison.

When IPCT Journal readers were asked what their primary criteria when choosing articles to retrieve (see Q9), it is not surprising that "interest in the topic" was mentioned by the largest number of respondents (68.2 percent). This was followed by 25.3% of the respondents stating "help with work," and 14.0% selecting "researching topic." Knowing the work areas of our subscribers will help us in choosing suitable articles for future issues.

Of the 240 IPCT Journal readers who responded and had a basis for judging, 84% characterized our articles as the same or better quality compared with those found in scholarly, refereed, print journals (see Q13). We believe another indication of usefulness and quality is whether IPCT Journal articles are cited by readers. When asked, 14.2% of the respondents to this question said they have cited the Journal's articles in their work (see Q18). Several noted that they have not as yet done so, but plan to in the future.. Finally, over half (53.2%) of the respondents said they had sent a copy of an article to a colleague (see Q19). This sharing of electronic resources is so easy via electronic mail that we know from personal experience that it is very common. When looking at criteria for choosing an article to retrieved 6.7% of the respondents said they knew someone who would like to read it (see Q9).

Sometimes the electronic environment is complex. For example, in completing this survey, Thomas Berner (1994) writes:

Two thirds (66.3%) of the respondents to this survey received their copy of the IPCT Journal Table of Contents from the IPCT-J distribution list (see Q7). Another 15.7% received their copy from the related IPCT-L discussion list.. The other LISTSERV discussion lists to which the Table of Contents is crossposted vary from issue to issue. Some crosspostings are chosen by the publisher to widen the exposure of IPCT Journal, and some crosspostings are made by readers. Surveys from postings to these other lists accounted for 9 percent of the responses. Additionally, respondents retrieved copies of the survey from a gopher (2.3%), an FTP site (1%), the bit.listserv.ipct-l Usenet newsgroup (1.8%), another person (2.1%) and other sources (1.8%).

We asked whether subscribers preferred to read IPCT Journal articles on-screen or to print them out on paper to read, and what that choice depended upon (see Q20). Well over half (58.4%) stated they read articles on-screen and print them out. This decision depends upon many things, but the following groupings are listed by a number of respondents in an open-ended question:

Each of these could be interesting avenues for follow-up research.

Finally, with regard to the technical aspects of IPCT Journal, we asked the readers in an open-ended question their preference for formats besides the current ASCII format. Over 100 readers responded that they would prefer HTML, PostScript, Word, and WordPerfect. Several readers noted, however, that if the articles for IPCT Journal are text only (i.e., non-graphics) then something like PostScript would not be preferred (see Q12).

Without a doubt, one of the biggest barriers to scholarly electronic publishing concerns promotion and tenure criteria and the committees that use them. While the manifest purpose of scholarly publishing is for scholars to disseminate their work to advance knowledge, most scholars want to do so in a way that will derive the most benefits for them, including advancement in their field and at their institution.

We asked our readers how electronic publishing was viewed by promotion and tenure committees at their institution when compared to publications in print journals (see Q17). Of those persons who are in higher education and believe they know enough to make this comparison, 33.8% said the same points are given regardless of the format the scholarly article. This seems to us to be a positive indication of a growing acceptance at this point in the history of scholarly ejournals. Not surprisingly, the respondents to this ejournal overwhelmingly (86.8%) reported that they believe articles in ejournals should count the same as in print journals.

While the entire population of IPCT Journal and IPCT-L subscribers (and an unknown number of other persons) received at least one copy of the Table of Contents containing this survey, only 16% of the 1585 IPCT-J subscribers responded to the survey. There are several groups of readers who would not be reached by the survey included in the Table of Contents, including those reading individual articles distributed by libraries and those shared by colleagues. We can liken the reception of email surveys in at least one way to that of many postal mail surveys, with the recipients using their "delete" key instead of their wastepaper baskets.

Even though respondents had the option to fax or use the post office to mail in their completed surveys, all but a dozen or so persons chose to return them via email. Because responding to the survey online required some expertise in using email, there may be a bias in responses toward persons with such skills. Because receipt was predicated, for the most part, on access to a computer, email and the Internet, this may also tend to push the respondents to the higher end of the educational and socioeconomic continuums.

Given these limitations, it may be that the respondents do not represent IPCT Journal readers as a whole. Therefore, caution should be used when interpreting or generalizing based on these results.

There were several questions that, in hindsight, should be modified for clarity, and also several questions that should have been asked, but were not. For instance, asking the readers how the reality of the journal met their preconceptions (see Q14 and Q15) should have lead to a separate question: "how satisfied are you with the journal regardless of your preconceptions." Several readers mentioned which types of articles they would like to see more often in IPCT Journal, and this would have been an useful question, too. Additionally, it would have been beneficial to ask the readers which journals they see as competition to IPCT Journal.

The readers were asked their preferred format for obtaining the IPCT Journal text (see Q11). Unfortunately, we failed to include "email/LISTSERV/ASCII" (i.e., LISTSERV, listproc, majordomo, mailserv or UNIX mailing list) in the choices. While IPCT Journal is currently available via Listserv, gopher and FTP, to not include the most prevalent media is an oversight that casts doubts on the responses to this question. It will be interesting to track the preference for WWW (and Mosaic) over time.

There is no way of knowing how well the respondents to the IPCT Journal survey actually represent all IPCT Journal readers. Persons returning the questionnaire are a self-selected sample. Still, given the number of responses, the geographic distribution and other characteristics of the respondents, we believe the data collected is of value. It is clear, however, that there is no single academic discipline that IPCT Journal serves.

From the number of times the respondents said they would retrieve, and have retrieved articles from past issues, and what others have said above, we hypothesize that ejournals may have substantially higher readership per article than print journals, and we await the technological advances that will allow us a relatively accurate count of electronic article retrievals.

Readers’ perceptions of the credibility of the articles appearing in IPCT Journal is indicated by the use of IPCT Journal citations in their own work (14.2%), choosing articles on the basis of their usefulness in their own work (25.3%), for others (8%), and their willingness to pass articles along (53.2%). Most encouraging from the publisher's perspective is readers’ perception of the equal or higher quality (49.2%) of IPCT Journal articles compared with refereed articles in print journals.

Many respondents wrote in that ASCII via email was their preferred format for initial delivery. That is understandable as it can be imported into any word processor. Gopher was the most popular form (40.9%) of storage/retrieval. As user-interfaces improve it will be interesting to track what changes occur in preferences for technical access. We surmise that preferences may tip in favor of retrieval from the World Wide Web, as it appears to be enjoying an explosive growth and is easy to use, although it does require a high speed internet connection for best performance/

Permanence is seen by us as an issue largely in the hands of ejournal publishers and it was not referenced in this survey. Acceptance of ejournals by this readership is indicated by 85.4% of respondents subscribing to more than one ejournal. Acceptance by promotion and tenure committees is still largely problematical with 24.9% of respondents saying that retention, promotion and tenure committees at their institutions awarded less or no points for published articles in ejournals and only 12.7% saying they gave equal or more weight. (Sixty-two and three tenths percent of respondents were either not in higher education or didn't know the weight their institutional promotion and tenure committees gave to ejournals.)

The respondents to this survey are largely middle-aged males, skilled with both computers and email and working largely in academic settings. "Generally help with my work," was chosen by 25.3% of the respondents as the most frequent criterion used in selecting articles to retrieve, so it would behoove ejournal publishers to determine the kind of work their subscribers do and actively solicit articles that would pertain to those areas.

A readership census of DEOSNEWS, an ejournal distributing one article a month to its subscription list of over 2400 persons interested in distance education is currently underway. The census is being conducted by the authors and uses a modified version of the IPCT Journal survey. This will allow many points of comparison. We hope to conduct several more readership surveys of ejournals in different fields so we might build a generalizable model of ejournal readership and readers.

We have anecdotal evidence that indicates some differences between electronic and paper based surveys which will be interesting to investigate, given the increasingly wide spread of access to electronic mail. In general, the information about our readers gained from this first readership survey, and the stability and slow increase in our subscription list encourages us to continue to publish the journal and to seek out articles that are more relevant to the needs of our audience.

FOOTNOTES

This was set up to be viewed on a terminal with a display 80 characters wide and 22 lines long. We have use monospaced type to preserve the formatting and the original onscreen appearance. The number and frequencies of response have been included.

READERS' SURVEY - IPCT JOURNAL

Please place an X or other character beside your answers below. The term "electronic journals" (ejournals) for this survey is NOT a synonym for other forms of electronic communication like discussion lists. We are referring here to electronic entities, like IPCT-J, which publish full-length articles.

1. Using computers, I consider myself a(n): (N=394)

1.5% beginner 19.3% intermediate

45.9% advanced intermediate 33.2% expert

2. Using email, I consider myself a(n): (N=394)

2.5% beginner 21.3% intermediate

44.4% advanced intermediate 31.7% expert

3. My age group: .3% <20 8.9% 20-29 29.9% 30-39

(N=394) 44.4% 40-49 14% 50-59 2.5% >59

4. My gender: (N=391) 29.2% Female 70.8% Male

5. Besides the IPCT Journal, how many electronic journals

do you subscribe to/receive? (N=390)

14.6% none 35.5% 1-2 34.4% 3-5 15.1% more than 5

6. Where is the email address at which you received this issue of IPCT-J?

City/State: Country: 71.6% US 28.4% Other

7. Do you (primarily) get your copy of the IPCT Journal Table of Contents from (check one only): (n=389)

66.3% IPCT-J@GUVM list

15.7% IPCT-L@GUVM discussion list

2.3% gopher

1.0% ftp site

1.8% Newsgroup (Usenet)

9.0% another discussion list online--which one?

2.1% forwarded to you by another person

_1.8% other source __________________________

8. How many articles from this issue of IPCT Journal are you likely to request in full text? (N=386)

21% none 51% 1 or 2 17.9% 3 or more 10.1% all of them

If not all, which article(s) will you request?

(Please list the titles or authors):

9. What criteria do you (PRIMARILY) use to choose the articles you read: (please check only one)

(Note: despite the instructions many persons checked more than one item)

57.4% I am interested in (topic)

6.7% I know someone who would like to read this

11.7% I am researching this myself

21.3% This will help my work generally

2.8% Other -- Please describe briefly

______________________

Any additional comments?

10. If you have received previous IPCT Journal issues, how many articles did you generally request from each issue? (N=357)

This is my 25.5% 1st issue

14.8% none 40.3% 1 or 2

7.3% 3 or more 5.0% all of them

7.0% it depends on ____________________

11. If I had my choice, I would prefer that IPCT Journal text could be obtained in which format (check only ONE): (N=342)

2.9% on CD ROM

1.8% on Floppy disk

40.9% Gopher

24.9% WWW

17.5% FTP site

12.0% other _______________________________

12. Currently, IPCT Journal is published in ASCII only. Would you prefer it to be in some other format as well, (e.g.,PostScript)? (N=383)

72.6% No 27.4% Yes

If yes, which one? _____________________

13. With regard to quality of the articles you have read, how would you characterize IPCT Journal articles compared with scholarly refereed print journals? (N=385)

5.1% better quality

47.0% same quality

9.1% somewhat less quality

1.0% much less quality

37.7% I have no basis to answer this question

14. What did you expect IPCT Journal to be when you subscribed (use as much space as you wish)?

15. How well does IPCT Journal meet your expectations in the question above (use as much space as you wish)?

16. Please tell us the general nature of the organization you work for: (N=389)

6.7% K-12 (Kindergarten through 12 years of schooling)

70.2% higher education

7.6% business, industry

5.1% government

10.3% other ______________________________________

Your field (e.g., Education, Computers, History):

Your specialty:

17. If you are in Higher Education, please respond:

I would characterize the behavior of administrators and members of committees who decide on retention, promotion and tenure at my college or university with regard to electronic publication of refereed journals such as IPCT Journal in the following way: (N=361)

12.7% an author receives THE SAME points for articles published in peer reviewed electronic journals as if that article was published in print

14.1% an author receives SOMEWHAT LESS points for articles published in peer reviewed electronic journals compared to if that same article was published in print

10.8% an author receives NO points for articles published in peer reviewed electronic journals compared to if that same article was published in print

24.9% not applicable (I am not at a college

university)

37.4% I don't know

Do you think that electronic journals like IPCT-J should

be viewed the same as print journals for promotion and

tenure?) (N=301)

86.0% Yes 13.0% No

18. Have you ever cited an IPCT Journal article in your own work? (N=358)

14.2% Yes 85.5% No (.3% Don’t Know)

19. Have you ever sent an IPCT Journal article to a colleague? (N=358)

53.2% Yes 46.5% No

20. Do you print out IPCT Journal articles or read them on-

screen? N=363)

23.4% on screen

18.2% printed out

58.4% both, it depends on _________________________________

THANK YOU FOR YOUR PARTICIPATION.