450 completed questionnaires were received from 450 sites. 81
respondents identified themselves as NIH grantees. Figure 1 (p. 14)
displays graphically the mean response for each item accompanied by bars
showing 1S.D. and the effective sample size for each item. The
left panel pertains to the existing program and the right panel relates
to planned features not yet implemented.
User Profile (items 1-7, 13, 14)
An overwhelming majority of VMD users are affiliated with
academic institutions (84%) and use VMD for research (78%). 19% of the
respondents reported to be funded by NIH. Almost all VMD users run the
program on Linux and IRIX. About 60% of the respondents would also like
to be able to use VMD on Windows 9x and Windows NT. Most of our users
first heard of VMD via the web (53%) or from friends (26.5%), and they
clearly prefer to be informed of VMD news by Email (57.9%) or web
announcements (38.8%). The respondents reported an average of 2.3 users
Evaluation of Present Program (items 8a-8d, 12)
On a scale from 1 (= strongly disagree) to 5 (= strongly
agree), the mean satisfaction score of VMD was 3.84 with an S.D. of .79.
The overall satisfaction is highly correlated with the mean response to
all evaluative items (r=0.55, p<.05). The joint distribution of these
two variables is shown in Figure 2 (p. 15).
The respondents reported that they use VMD because it:
Having a free graphics program is considered an obvious
advantage, alongside strong technical and GUI qualities. The low variance
produced by the respondents indicates a high level of agreement on all
Expected Usefulness of Future Features and of Interfaces to Other Tools
The expected usefulness of 7 future features of VMD was rated
on a scale from 1 (=unimportant) to 5 (=very important).
The lowest rated features were:
The top rated features were:
The respondents felt that they would benefit most from using
VMD with four of the tools listed: Amber (118), Charmm (111), XPLOR (94)
and Insight (85).
The Quality of VMD Support
The quality of VMD support was assessed on a scale from 1 (=
strongly disagree) to 5 (= strongly agree).
The users agreed that:
The clarity of the documentation and the instructiveness of
the web pages were ranked significantly higher than the other support
aspects measured. Here, again, the low variance produced by the
respondents indicates a high level of agreement on all items.
We performed several comparative analyses between various groups. In most cases we did not uncover any significant differences. There are only two noticeable exceptions. As shown in Figure 3 (p. 16) NIH-funded users ranked their overall satisfaction, as well as satisfaction with developers' responsiveness and the extent to which VMD meets their needs, significantly higher than non-NIH funded users. Figure 4 (p. 17) indicates that with a single exception academic users report a higher level of satisfaction with VMD, and on eight specific items this difference is statistically significant.