This page hosted by
Get your own Free Home Page
Cybermind is a mailing list, which was founded in June/July 1994 by Alan Sondheim and Michael Current, to discuss issues relating to the Philosophy, Psychology and Sociology of Cyberspace. Cybermind is an ideal place to study Internet groupings for the simple reason that many of the posts have always been off topic and information about the offline lives of members has always been available without causing too much disruption to the list. It provides constant illustrations of its subject matter.
Cybermind has always been a high volume list, on occasions reaching over 100 mails per day, on average it probably sees about 30-40 mails a day. During the first week of its existence it probably attracted over 300 subscribers, of whom 60 or so posted. Like most Internet groupings, the majority of posts came from a small number of posters - about 12 people were responsible for over half the posts. Six of these twelve most prolific posters were women. In the First month it appears that 25% of the active population was female and made 32% of the posts. Burstein & Kline, quote a Georgia Institute of Technology report from July 1995, which claimed that only 15% of Internet users were female (1995: 102). Almost 60% of the women subscribing to Cybermind at this time wrote from addresses which terminated in .ac or .edu which indicates that most were writing from academic institutions. This relatively high visibility of women lead fido, one of the more prolific female posters, to write (3 Jul.94): "One of the things I've found really refreshing about this list -- and this will be the first and probably the last time I hitch my post to a fe/male observation where I equate such distinctions with something positive -- is the number of female names attached to messages. I don't think I have ever been on a list with so many (presumed) female voices". This presence of women remained high at least until 2000 [because I have not done the third survey]
Population
Three population surveys were taken of the list [[i'm anticipating here not having done the third]], and the gender of respondent was compared to the assignable gender of those subscribing who did not respond. Clearly it was not possible to assign 'real gender' to most of those people who were subscribed but had rarely posted. For the first listing over 50% of the list was unassignable and just over 30% was similarly unassignable for the second listing. Comparing the results for both surveys of the population which was assignable a gender:
Table 1:
Responses 'Genderable' Responses 'Genderable' Responses
to pop. of to pop. of to
Survey 1 List 1 Survey 2 List 2 Survey 3
Female 26% 27% 36% 26%
Male 74% 73% 64% 74%
The average age of the women in the first survey was 38 and all but one were over 30, while in the second survey 5 women were under 30 and the age distribution of women was similar to that of the general list population - returning the same median and average. This change could represent a sample variant, but it could also suggest that a bias (whether of obstacle or inclination) against younger women using computers and the net had declined in the period between surveys.
The increase in the sample size of the second survey (from 44 to 51) was entirely due to an increase in the number of responses by women from 10 to 18.
Given the overwhelming proportion of males to females on the list, it is perhaps inevitable that the majority of posts will originate from males. However the following table, while not implying the list is not predominantly controlled by men, shows that women on this list are not silenced. The proportion of posts made by women is pretty much in line with the proportion of women in the list population, including the most prolific sections of that population.
Table 2: Presence of Women month (1995) A B C D E F G H I Mar 28 23% 269 21% 25% 194 21% 17 23% June 25 20% 463 23% 20% 303 22% 11 17% Sept 25 27% 507 30% 27% 294 26% 08 16% Dec 29 26% 342 28% 27% 205 26% 16 26% A = No. of Women posting B= percentage of active population that is female C = No. of posts made by women D= % of total posts made by women E = percentage of women in the 20% most prolific posters F = No. of posts made by such women G = Percentage of the posts made by the 20% most prolific posters made by women. H = No of women making 5 or less posts I = percentage of those making 5 or less posts who are women
I investigated a possible correlation between gender and response rates, to see if males responded primarily to other males, or whether female posters were ignored and so on. This considered 900 consecutive posts within March 1995. Of these 900 posts it was possible to classify 798 of them as ether male responses to males, male responses to females, female responses to males, female responses to females, males starting a thread, or females starting a thread. The remaining posts were either by people of unclear gender; posts in which it was not obvious to whom the poster was replying; and posts addressed generally to the list but not initiating a new thread.
In this final sample of 798 posts, 598 (or 75%) were made by males and 200 (or 25%) were made by females. Distribution of responses were as follows:
Table 3: Response Rates and Gender
M to M M to F F to M F to F M start F start
number 386 126 123 48 86 29
percent 48% 16% 15% 6% 11% 4%
of total.
Clearly 63% of these posts were responses to males, and 22% were responses to females.
If we take the percentage of these responses as a fraction of all posts by that gender - for example, the percentage of all female posts to males as a fraction of all female posts and tabulate them, then:
Table 4:
M to M M to F F to M F to F M start F start
65% 21% 61% 24% 14% 15%
There appears to be little significant difference between the response rates of males and females to each other, though males responded slightly less to females than females did, and females responded slightly less to males than males did. Both genders tried to initiate threads equally.
Comparing the number of posts by a gender and the number of responses to people of that gender, then 85% of male posts were responded to, and 87% of female posts were responded to.
In order to investigate whether the posts sent in by males were longer than those sent in by females I looked at one week between the 11th and 18th of April 1995. After allocating posts by gender, I counted the original lines of text, ignoring text which had been quoted. I also ignored a few posts which were primarily lists of Internet sites, or computer programs. All together the count involved 264 posts, and some 3,549 lines of text. There were 200 posts from identified males, and 64 posts from identified females. Obviously just over 24% of posts came from women, which is slightly less than normal for that period. The males wrote 2686 lines of text, for an average of 13.43 lines of text per post, and the females wrote 863 lines of text for an average of 13.48 lines of text per post. Though further counting throughout the year, and looking at different circumstances, might turn up a different result, this does suggest that average length of posts is reasonably similar between genders on this list.
These rough figures suggest that on Cybermind there was little statistical difference in people's responses to others by gender. The question of type of response (i.e. whether responses to females are more hostile or dismissive than responses to males) has not been investigated due to my difficulties of making such evaluations.
The question arises as to whether this kind of gender equity on the list has changed over the years. Clearly it is impractical to carry out a continuous investigation of the matter, but given the appearance of sexual differentiation which appeared in some of the discussions made in early 2001 after the gender project was announced I carried out a further count.