|andrew.treloar.net||Andrew Treloar's personal website|
Respondents were asked to indicate which pieces of technology they had access to, either at home or at work. The intention was to determine what facilities an e-journal could assume from its readership. There is little point in adding sound or full-colour video if a majority of readers cannot benefit from this.
Table 7-10 shows a summary of the responses to this set of questions, ordered by the sequence of questions in the survey instrument. It is clear from table 7-10 that over two-thirds of the respondents have access to various components of technology to support multimedia (sound + CD-ROM + colour output) on their computers. A more detailed analysis of the raw data reveals that of the 999 who have access to personal computers, 756 also have access to CD-ROM drives, 710 have access to sound output and 899 have access to colour screens. The set of those respondents who have access to a personal computer with sound output, CD-ROM, and colour output is 641 of the 1038 total respondents or 61.75%
What therefore is the detailed technology picture arising from these surveys?
There was some confusion about what constituted a personal computer (PC). Some respondents indicated that they considered a Unix workstation as a PC, while others listed this in the Other section. Because of the difficulty in reinterpreting the data, the responses have been left as received.
Of the total pool of respondents, 96.24% have access to a personal computer (however they define that). There are no statistically significant correlations between this variable and either the Society-Print variable (which distinguishes between the different print survey subgroups) (p = 0.4090) or the Survey variable (which distinguishes between the email and print surveys) (p = 0.2061). In other words, the entire survey population can be treated as single group with respect to this variable.
Nearly three-quarters of the respondents (73.12%) have access to a CD-ROM drive. As with the previous variable, there are no statistically significant correlations with either Society-Print (p=0.2774) or Survey (p=0.6875).
Nearly as many respondents (87.09%) have access to colour screens as have PC's. There are no statistically significant correlations with Society-Print (p=0.3455). That is, the differences in responses between print subgroups probably arise by chance. There is, however, an correlation between Colour and Survey (p=0.0045). In other words, the chance is less than 5% that the differences in access to colour output between the email and print groups is due to chance. The email group shows slightly greater but statistically significant access to colour output than would be expected by chance (and the print group shows correspondingly less access to colour output). Table 7-11 shows the frequency distribution and expected versus observed values.
Nearly three-fifths of the respondents (59.44%) had access to a direct network connection to the Internet. Table 7-12 shows the frequency distribution for the answers to this question, broken down by Society. If one performs a contingency analysis then there is a statistically significant difference (p<0.0001) between the responses from the different subgroups. One way to examine these differences is via the observed and expected frequencies. Table 7-12 also shows these values for this variable.
Another way to see the relative contributions of each cell is the post-hoc cell contributions table. Table 7-13 shows the contributions of the different subgroups to the contingency table chi-square p-value. A value of 1.96 is significant at the 0.05 significance level, and thus all of the societies show significant differences between expected and observed values (assuming the hypothesis of independence). The APA respondents show much less access than expected with 246 negative observed answers versus 197.12 expected. The APS respondents demonstrate the same picture: 57 negative observed answers versus 32.29 expected. The BPS respondents on the other hand show significantly more access to a direct network connection than expected with 89 positive observed answers, as against 76.68 expected. The largest post-hoc contribution comes from the email survey respondents who have a contribution of 7.874 and an observed positive count of 258 versus an expected value of 199.72. The email survey respondents had the greatest access to direct network connections, followed in order by the BPS, APA and then APS respondents.
Just over three-fifths of the respondents (60.308%) had access to a modem connection to the Internet. As with the network question, there is a statistically significant difference (p<0.0001) between the responses from the different subgroups. The post-hoc cell contributions of each of the different subgroups to the contingency table chi-square p-value are significant at least at the 0.05 probability level. Table 7-14 shows observed and expected frequencies broken down by Society. The APA respondents show more access to a modem connection than expected. The APS respondents show less access (as with the network question). The BPS respondents also show significantly less access than expected. The email survey respondents again have the highest proportion of positive answers (although the APA respondents are very close).
Last modified: Monday, 11-Dec-2017 14:39:25 AEDT
© Andrew Treloar, 2001. * http://andrew.treloar.net/ * firstname.lastname@example.org