The claim that a medium has provided the basis for the formation of a large community that is important to its members is a fairly exceptional one. It is generally reserved for highly successful and widely used media. If computer conferencing and other forms of computer mediated communication can form the basis of a new electronic community, it is then reasonable to expect fairly high levels of success as measured in its use relative to use of other media. Given the apparent formation of a moderately strong electronic community on IBMPC, moreover, it is reasonable to expect fairly high levels of use of computer media, including electronic mail and computer conferencing, among IBMPC participants.
A set of questions in the 1988 IBMPC user survey directly addresses this expectation, with respondents asked to estimate the number of hours they spent using a variety of media. In addition to computer conferencing and non-conference computer mediated communication, the survey collected information concerning respondents use of lectures, correspondence, the telephone, group meetings, radio, television, publications, and face to face interaction. These questions extend a second set of questions (asked in both surveys and reported on in the appendix on demographics) which asked respondents to estimate the amount of time they spent performing various computer based tasks.
It is possible to rank the media assessed in the 1988 survey in terms of their relative use. Based on tests of statistical significance, the ten media categories break up into six distinct sets. Four media stand alone, with levels of use that are significantly different than all others. Six media cluster into two fairly well defined groups. A summary of the results is shown in the following table:
| Usage Level by Percentage | ||||||||||
| Measures of Media Use | 35 hours or more | 17 to 35 hours | 6 to 16 hours | 1 to 5 hours | Less than 1 hour | N | Mean | St. Dev. | ||
| Lecture | 0.0 | 0.0 | 1.7 | 22.0 | 76.3 | 177 | 1.25 | 0.47 | ||
| Correspondence | 0.0 | 0.0 | 2.8 | 31.1 | 66.1 | 177 | 1.37 | 0.54 | ||
| Telephone | 0.6 | 2.8 | 13.6 | 55.9 | 27.1 | 177 | 1.94 | 0.75 | ||
| Group Meeting | 0.0 | 0.6 | 24.9 | 59.9 | 14.7 | 177 | 2.11 | 0.64 | ||
| Conferencing | 0.0 | 2.3 | 24.9 | 57.6 | 15.3 | 177 | 2.14 | 0.69 | ||
| Radio Hours | 1.7 | 5.1 | 24.9 | 50.3 | 18.1 | 177 | 2.22 | 0.86 | ||
| Electronic Mail | 1.1 | 12.4 | 27.1 | 50.8 | 8.5 | 177 | 2.47 | 0.86 | ||
| Television | 1.7 | 7.3 | 43.5 | 33.3 | 14.1 | 177 | 2.49 | 0.89 | ||
| Publication | 0.6 | 6.2 | 53.1 | 37.9 | 2.3 | 177 | 2.65 | 0.66 | ||
| Interaction | 13.0 | 32.8 | 42.9 | 10.7 | 0.6 | 177 | 3.47 | 0.87 | ||
The lecture was the least used media among IBMPC users, with less than a quarter of respondents listening to more than an hour of lecture a week. This should not, perhaps, be surprising. Most peoples experience of lectures curtails to almost nothing when their schooling is complete.
Respondents indicate spending significantly (t=2.18, p=.03) more time on non-electronic correspondence than they do in lectures, but not much more. Only a third of respondents claim to spend more than an hour a week on their correspondence. This is, perhaps, a more surprising finding, however, as preparation of only one or two memos a week could easily eat more than an hour a week, especially when combined with the personal mail received at home that one must, at least, read. It seems obvious that if this result is correct, that it represents a substantial reduction in overall use of hard copy mail.
As no questions on non-electronic media were asked in the 1986 survey, it is not possible to document an actual decline in the use of non-electronic correspondence. There has, however, been a significant increase in the use of electronic correspondence between 1986 and 1988 (reported on later in this chapter), and that may well have been accompanied by a correspondingly significant decline in the use of hard copy mail.
IBMPC participants use of the telephone is significantly greater than either correspondence (t=8.47, p<.0001) or lectures (t=10.57, p<.0001), and significantly less than any other media (minimum t=2.51, p=.0129). The strong median for telephone use is 1 to 5 hours. 56% of respondents indicated this level of use. Fully a quarter of respondents, however, indicated less than an hour of telephone use a week, and only 20% claimed to use the phone six or more hours each week.
This result is probably about right. Most people probably do use the phone for between 1 and 5 hours a week. There are strong indications that people in IBM have reduced their telephone use somewhat based on their use of electronic communication. Employee business cards increasingly include both an employees telephone number and the node and userid information required to send electronic mail, and many employees express an active preference for electronic correspondence. Indeed, many of the telephone numbers listed in IBMPC's WHOSWHO FORUM (a sort of voluntary directory of IBMPC participants) are accompanied by admonishments to use EMAIL, including "VNET preferred", "Please send EMAIL", "Don't Like Phones", "You will never reach me. EMAIL is better.", "I prefer EMAIL", and "Much prefer EMAIL instead. Not as disruptive.".
This preference for EMAIL over the telephone is clearly reflected in these media rankings, as claims for both computer conferencing and other computer-mediated communication (still to be reported) are significantly greater than those for telephone use. The reasons for this preference probably center on two factors. First, the telephone, when it rings, is an interruption which demands immediate attention that must be diverted from whatever one is working on. Second, the telephone is a synchronous medium. A connection can only be completed if both the caller and the callee are in the right place at the same time. The synchronization frequently fails, and even with the intercession of answering machines and messages, the result is often "telephone tag".
EMAIL, because it is asynchronous, non-interruptive, and fast, can frequently solve these problems. The message moves from sender to sendee almost as fast as a telephone connection might be made, but the sendee is able to look at and respond to the message at their convenience. The experience of exchanging electronic mail with others several times during a bout of telephone tag is a common one in IBM, and many people use electronic mail to set up mutually agreed to "telephone appointments". Hence while telephone use is still widespread within IBM, the nature of telephone use has probably changed, and the extent of telephone use probably reduced somewhat, as a result of the widespread use of electronic media.
Group meetings, Computer Conferencing, and Radio use cluster together in the 1988 survey as media that are used to a similar extent. Indeed, all have a strong median at 1 to 5 hours of use per week, with a range from 50% for radio to 60% for small group interaction. Roughly 15% of respondents claim to use each of these media for less than an hour a week. Roughly a quarter of respondents claim to use each of these media for 6 or more hours a week.
The rough parity of use experienced by these three media is interesting for at least two reasons:
It is interesting to note that while levels of use for radio and small group meetings probably remained fairly stable between 1986 and 1988, use of computer conferencing grew significantly. In both surveys, 57% of the survey respondents indicated that they spent between 1 and 5 hours a week on computer conferencing. This strong median masks, however, a significant shift (t=3.38, df=298; p<.0005) in computer conferencing usage between 1986 and 1988. 29% of the 1986 respondents spent less than an hour a week engaged in computer conferencing (14% spent 6 or more hours a week on conferencing). 27% of the 1988 respondents spent 6 or more hours a week on conferencing (only 15% spent less than an hour a week conferencing).
A second cluster of media, this one composed of electronic mail (non-conference communication), television, and publications (including both books and periodicals) sets a still higher standard of use. At least 40% of respondents claim to spend at least 6 hours a week with each of these media classifications. These clustered is centered by television, which is not significantly different in its estimates than electronic mail or publications. Electronic mail and publications bound television, on the low and high ends of usage, respectively, with publications used significantly more (t=2.22, df=175, p<..02) than electronic mail.
The relative success of television as indicated by its presence in this grouping is not, perhaps, unexpected. Fully 52% of respondents claim to watch television for at least 6 hours a week (the median category is 6 to 16 hours). Still, it is not as successful as one might expect, with only 9% of respondents claiming to watch more than 17 hours of television a week. This result, which indicates that the median IBMPC user watches roughly 2 hours of television a day, might be explained away as an underestimate. One notes, however, that 45% of the 1988 sample indicated that they spent at least 6 hours a week making personal use of a computer, most likely at home. This relatively high level of home computer use probably does reduce the amount of time available to watch television.
More surprising, perhaps, is the high level of use associated with publications, including newspapers, magazines and books. A full 70% of the sample indicated that they spent 6 or more hours a week reading various publications. This should not, perhaps, be unexpected, as this level of use can be attained with less than an hour a day, on average, devoted to reading the newspaper. Professional demands for reading documentation, sales literature, reference books, manuals, and journals only adds to this baseline.
Most surprising, however, is the relatively high elevations occupied by electronic mail. Although its median (56% in 1986 and 50% in 1988) level of use, 1-5 hours, is lower than either television or publications, the survey results clearly suggest that electronic mail has grown between 1986 and 1988. Respondents who communicate electronically for less than one hour a week decline from 23% in 1986 to 9% in 1988 while those spending six or more hours a week engaged in non-conferencing electronic communication expands from 21% in 1986 to 39% in 1988. The resulting shift is strongly significant (t=4.72; df=298; p<..00001).
Use of electronic mail has grown, between 1986 and 1988, to a position where its use is comparable, at least for IBMPC participants, to television, and only slightly behind publications. This is interesting, in part, because of the heavy push in the computer industry to move publications from paper to the computer screen. If this push is successful, it seems likely that electronic mail's climb to success is not over yet.
The most heavily used medium among those explored in these questions is face to face interaction. 89% of respondents claim to spend at least 6 hours a week interacting directly with others, and 46% claim to spend at least 17 hours a week doing so. This is hardly a surprising result. One can achieve the first of these levels by simply talking to other people at meals.
What is interesting, however, is a comparison of face to face interaction hours with the measures of computer use. While the survey contains no measure of total computer use, it does contain a measure of the number of hours spent using a computer for work related activities. This measure confounds any number of elements, and must be assumed to include hours spent engaged in computer conferencing and electronic mail. Even given these caveats, it is interesting to note that respondents to this survey spend significantly more time talking to their computers than they do talking to other people face to face.
One hesitates to interpret this last finding too strongly. The same group claims to interact with others significantly more than they watch television. Recent surveys of the U.S. population have indicated that the opposite is more generally true. One remembers, also, that the computer is, for many of these people, their livelihood. If the computer is regarded as a new form of pencil and paper, such statistics may not be unreasonable.
Taken as a whole, these results seem to indicate that computer-mediated communication systems, in the form of computer conferencing and electronic mail, have become very successful within IBM, at least among participants in the IBMPC Computer Conferencing facility. Not only has use of these media grown substantially between 1986 and 1988, but these media are now competitive, at least in terms of the hours users spend engaged in them, with other media that have traditionally been considered highly successful, including small group interaction, radio, television, and publications; media which few would dispute have had had a substantial effect in changing the human environment.
The level of success of computer conferencing and electronic mail would seem to indicate that: