Thesis --> Notes
[1] As identified by their administrators or supervisors. [return to text]

[2] It has become commonplace in academic usage to extend the use of the word “gender” to encompass the meaning of the word “sex.” However, the Oxford English Dictionary defines gender as “a euphemism for the sex of a human being, often intended to emphasize the social and cultural, as opposed to the biological, distinctions between the sexes.” In the interests of precision, I shall refer to biological differentiation as sex. [return to text]

[3] Not all forms of CMC are asynchronous. Any mode or medium of communication transmitted through a modern telecommunications system is handled by computers; in its broadest sense, CMC encompasses all forms of modern telecommunications. For the purpose of this work, however, I am ignoring synchronous forms of CMC such as UNIX's talk, Internet Relay Chat or “chat rooms” on commercial access providers, as well as audio, audiographic, or video conferencing. Rather, I shall focus on asynchronous forms of CMC such as electronic mail (including listservs), newsgroups, and groupware. [return to text]

[4] Whether these people are truly as influential is a different matter. One might imagine status differences between ideas that were presented FTF and those presented via e-mail, for example. [return to text]

[5] For our purposes, this form of synchronous communication is similar to FTF. [return to text]

[6] This is similar to the old argument between users of computers with command-line interfaces and those with graphical interfaces. They maintain, respectively, that one is traditional, more powerful, permits the user more control and requires a greater understanding of the computer whereas the other permits one to concentrate on the task at hand rather than the interface (Heid & Norton, 1989; Naiman, Dunn, McCallister & Kadyk, 1992; Norman, 1988; 1990). [return to text]

[7] Writing apprehension is used to describe a variety of behaviors that include a tendency to experience anxiety when faced with a writing situation, a tendency to avoid situations that require writing and a tendency to experience frustration and hesitation while actually writing (Daly, 1978; Daly & Miller, 1975a; 1975b). [return to text]

[8] Participants used Lotus Notes version 3.30, mostly on Intel-based machines using Microsoft Windows. I used the Macintosh version, also 3.30. [return to text]

[9] The other courses were Systems Auditing and Information Security. These were taught after the Systems Analysis and Database Management courses. [return to text]

[10] The instructions for the transcriptionists are reproduced in Appendix B. [return to text]

[11] Since the author of each CMC contribution could be identified by the file's header information (see Figure 6), this information was removed from each file and placed into a separate database, the metadatabase (see Figure 8). The metadatabase contained the original identifying information such as the author's sex, classroom role, the course, and term. [return to text]

[12] The coder's initials. [return to text]

[13] AppleScript was released with System 7 Pro in 1993; it is part of the Macintosh operating system. [return to text]

[14] The term “macro” is derived from the idea that several “micro” commands are components of macro commands (Naiman et al., 1992). [return to text]

[15] Applications which incorporate macros include Claris's FileMaker Pro, Adobe's PageMaker, Software Venture's MicroPhone Pro, Microsoft's Excel, and WordPerfect Corporation's WordPerfect for Macintosh. [return to text]

[16] Although other tools for the manipulation of Apple events are available, such as UserLand Software's Frontier or CE Software's QuicKeys, I have used AppleScript because it is included with every new Macintosh. [return to text]

[17] Modem connections slow Apple events significantly, however, so one would only want to do this over at least a 9,600-baud connection (DiNucci et al., 1994). [return to text]

[18] AppleScript requires a Macintosh configured with a 68000 or greater processor (Apple Computer, 1995) (in other words, any Macintosh or processor capable of running the Mac OS), a minimum of four megabytes of RAM, and System 7.0 or later. According to the MacScripting Frequently Asked Questions list (Terry, 1995), however, at least a 68020 processor is required to run the latest version of AppleScript. This would exclude the 128K, 512K, 512Ke, Plus, SE, Portable, Classic, PowerBook 100, and the Outbound 2000 (DiNucci et al., 1994; Naiman et al., 1992). From a more pragmatic point of view, AppleScript requires enough RAM or virtual memory to open the operating system and all applications which are to be used. If one is running System 7.X on a Macintosh with 4MB of RAM, there isn't a lot of RAM left for running applications, and virtual memory is only available to Macintoshes with 68030 processors or PMMU chips (Naiman et al., 1992). Unless adapted, most of the AppleScripts described in this work will not work on a Macintosh containing less than a 68030 processor. [return to text]

[19] Although I'd read about Apple events from the computer periodicals before the release of System 7, I had little experience with them until 1993, when I worked as a technician at a software company called Quark. Part of my job at Quark, which is best known for a page layout application called XPress, was to test the new version of XPress before it shipped. The new version of XPress was Apple event-aware, and part of my work included testing the scripts that came with the package. [return to text]

[20] Estimates based on a Quadra 700, upon which I wrote the initial scripts. [return to text]

[21] The initial version of Word 6, with which I first wrote AppleScripts, was slower than Word 5.1 (Markoff, 1995). Critics described its performance as “average” (Norr, 1995), “disappointing” (Markoff, 1994), “incredibly sluggish” (Seymour, 1995), and “wretched”. “It's as if a car company went bonkers with moon roofs and audio systems but produced a car whose top speed was 45 miles per hour” (Levy, 1995). In response to “cries of anguish from many in the Macintosh community” (Markoff, 1994), Microsoft released a bug fix in March of 1995 to address some performance issues, but it hasn't helped very much. Word 6 still “moves like a giant northwestern [USA] slug” (Rizzo, 1996b). [return to text]

[22] Word 6.0.1 for the Power Macintosh suggests a RAM allocation of 6.8MB, compared to Word 5.1's modest 1MB. In effect, this means that a base-configuration Power Macintosh cannot run Word 6 unless virtual memory is used, which slows the performance of any application to a crawl. Furthermore, Word 6 requires nearly 25MB of hard drive space for a normal installation. [return to text]

[23] Word 6's violations of Apple's interface guidelines have resulted in a product that is “harder to learn, more difficult to manage, and much less pleasant to use than its predecessors” (Levy, 1995). Its interface has been dismissed as “Windows-like” (Seymour, 1995) with an “un-Macintosh look and feel” (Norr, 1995) that is “a complete and utter disgrace” (Ihnatko, 1995). Microsoft boasts that “Word for Windows and Word for Macintosh share the same file format, features, appearance, and documentation” (Word 6.0 manual, p. xiii), and one glance at the screen shows that the interface has indeed sunk to the lowest common denominator. Ihnatko (1995) describes it well: “This is not Word for Macintosh. This is undeniably Word for Windows for Macintosh.” Word 6 has become such a laughingstock in the Macintosh world that a recent review of accounting software (Gallagher, 1996) described the genre as “more fun than Word 6.” [return to text]

[24] MacLinkPlus is an excellent conversion application that as a stand-alone utility permits conversion between nearly every popular Apple II, Macintosh, DOS, Windows, or UNIX application. In addition, MacLink extends the capabilities of an application to save in these various formats. [return to text]

[25] Although changing that initial character to a larger point size would prevent me from seeing other characters on that first line of the paragraph, the resultant change in leading would also have made it more difficult to see whether a paragraph had been missed. [return to text]

[26] Selecting male or female students by their class representation follows from my assumption that all students are equally likely to contribute to the course discussions (see Appendix J). [return to text]

[27] Although one may presume safely that longer passages were from instructors. [return to text]

[28] Created with the shift-option-hyphen keys. [return to text]

[29] Created with the option-semicolon keys. [return to text]

[30] My particular favorite, however, was that in the regional accent of Rhode Island (where my transcription was performed), one says “card,” as in playing cards, the same way one would say “cod” as in fish. As a result, references to E. F. Cod, the father of relational databases, were transcribed as “E. F. Card.” In a similar manner, references to “FoxPro,” a popular database, were transcribed as “Foxboro,” a nearby city in Massachusetts. [return to text]

[31] To be tidy, I also replaced the closing “])” with “]”. [return to text]

[32] This extra step was not necessary with the FTF records. However, the CMC to Database script uses string variables that are limited in the number of characters they can handle. For a more complete description of this problem, see Appendices F and G. [return to text]

[33] A line is a 4.5-inch segment of transcription in elite (12 characters per inch) type (p. 255); 54 or fewer characters. [return to text]

[34] I selected the nearest binder containing the printouts and opened it to a middle page. [return to text]

[35] Word 6.0.1's Word Count and Replace functions found 6,179 uhs and 1,521 ums in the 527,689 words of the FTF sentences. [return to text]

[36] Used after the speaker misspoke. [return to text]

[37] If the 8,782 sentences generated in the CMC courses were spoken at the same rate as the 43,299 sentences uttered in the 108 hours of FTF class time, this would take about 22 hours. [return to text]

[38] Assuming that a class lesson is 50 minutes in length, Bellack, et al.'s data were 50 minutes x 4 class lessons x 15 classes = 3000 minutes = 50 hours. Making the same assumption about class length, Sinclair & Coulthard's (1975) data were 50 minutes x 6 lessons = 5 hours. [return to text]

[39] The combined total of 99% is due to rounding error. [return to text]

[40] The combined total of 101% of this row is due to rounding error. [return to text]

[41] The first line of every CMC contribution was a subject line, each of which was coded Organising/Fact-Stating/Content. Of a total of 572 CMC files resulting in 8,782 sentences, 6.5% of all CMC sentences were subject lines. As a result, these three categories should be about 6.5% higher for CMC than for FTF interactions. [return to text]

[42] Although imperfections exist in written interactions (Finnegan, 1988; Graddol, Cheshire & Swann, 1994), one has a greater opportunity to correct them. [return to text]

[43] I imagine that this sort of thing will decrease in the future as more people grow up computer literate. At the present time, this is akin to having to teach a student how to use a pen to complete his or her homework. [return to text]

[44] MacUser [UK], 21 June 1996. [return to text]

[45] At the writing of this work, Apple are revising their operating system software to take full advantage of OpenDoc. [return to text]


hillman@cantab.net1997