OSS Watch Survey 2006: Report

1.1. The OSS Watch 2006 Survey

During February and March 2006, OSS Watch conducted a survey¹ of UK Higher Education (HE) and Further Education (FE) institutions, looking at their attitudes and policies towards open source software (OSS). This was in many ways a repeat of a similar exercise that OSS Watch performed in October 2003.

Improvements in the way in which the 2006 survey was performed resulted in two main benefits:

• By sending the survey's URL to the ICT director of the institution, the survey was more likely to have been completed by the person who is most able to provide answers about ICT for the whole institution.
• The survey was completed by more people: in 2003, responses (from those replying on behalf on their institution) were obtained from only 6% of UK HE and FE institutions; in 2006, a reply from 18% of institutions was obtained.

Whenever possible, comparisons have been made with the results of the 2003 survey. However caution has to be adopted in doing this since the roles and responsibilities of those who answered the 2003 survey are not equivalent.

1.2. Key findings

Institutional policies

Although only 25% of institutions mention OSS in an institutional policy, in practice 77% of institutions consider OSS when procuring software.

Contributing to OSS

16% of institutions do not know if their staff submit patches or code to OSS projects. Of those institutions where the ICT manager does know, 86% do not submit whereas 14% do. For those that submit patches or code, the regulations concerning this were varied:

• 7 respondents said it is the working practice of staff to supply patches;
• 6 respondents said that they do not know what the regulations are in regard to supplying patches;
• 5 respondents said that staff are supplying patches in their own time;
• 1 respondent said that this is specified in the employees's contract.

Use of the Moodle VLE

56% of FEs use Moodle as a Virtual Learning Environment.

Content Management Systems

There is no commonly accepted software for Content Management Systems (CMSs): the range of software being used for a CMS is wider than for other applications. In all, 69 institutions answered the CMS question giving 29 different answers.

Software for wikis and blogs

Although recently there has been a lot of discussion about the use of wikis and blogs, of the 23 institutions that answered the question about wikis, half of them said that their institution did not use wikis. Similar figures also apply for blogs.

Software mix on servers

Most institutions (69%) have deployed and will continue to deploy OSS on their servers. Generally, the software on servers is a mix of OSS and proprietary software (PS). The use of OSS is most common for database servers (used by 62% of institutions), web servers (59%) and operating systems (56%).

Reasons for choosing software on servers

When choosing PS for servers, the responses show that there is no one reason or a combination of reasons that leads to doing this. However, the reasons for using OSS are specific: saving on total costs of ownership, lower likelihood of getting locked in to a solution from a specific vendor, and better interoperability with other products.

Support of OSS

Most institutions that use OSS on their servers rely on in-house support for the OSS.

Software mix on desktops

In contrast to servers, the use of OSS on desktops is far less common. 47% of institutions indicated that, on desktops, they have not used OSS in the past and will not use it in the future. However, it is increasingly common to use both OSS and PS on desktops (42% do this).

Commonly deployed desktop software

Microsoft Office and Internet Explorer are deployed by all institutions on most desktops. Other commonly deployed applications are Microsoft Outlook (82%) and Mozilla/Firefox (68%). The latter's use is now considerably higher than in 2003.

Reasons for choosing software on desktops

The responses indicate that, saving on the total cost of ownership is, for desktops as for servers, the most important reason to use OSS on desktops. This was also indicated as the most important reason in the 2003 survey. Being locked in was not mentioned as an issue in 2003, but in 2006 the majority (73%) indicated that they chose OSS for that reason either on their desktops or on their servers or on both. Ideology is not an important factor for respondents, although a third of them indicated that it played a role.

1.3. Conclusion

A positive picture of the use of OSS emerges in both HEs and FEs. Although there are considerable differences between the two types of institutions, in general OSS is used more often than in 2003 and institutions have higher levels of skills and experience of OSS compared to 2003. This survey shows that it is likely that, in the future, use of OSS will continue and expand alongside the use of PS.

3.1. Questionnaire

The questionnaire³ for the 2006 survey was partly derived from the questionnaire used for the 2003 survey and from a questionnaire constructed for the FLOSSWorld project⁴ which ran in parallel to but was independent of this survey.

The questionnaire consisted of 45 questions organised into five topics:

1) Characteristics of the institution and the respondent

2) ICT policy regarding software and its deployment

3) Skills and awareness of ICT personnel in relation to software

4) Deployment of software on servers

5) Deployment of software on desktops

3.2. Response rate

One of the problems with the 2003 survey was its low response rate: of the 672 institutions contacted, only 63 replied. Of these 63 respondents, 22 answered for their department and only 41 answered for the whole institution (HE n=14; FE n=27). This is a response rate of 6%, which is very low.

It was important to improve the response rate for this survey since the suspicion was that only those institutions in which there was a higher interest in OSS would have replied, skewing the results. For the 2006 survey a list was composed containing the names of the individuals most likely to be in charge of ICT at each institution. These individuals, covering 637 institutions, were contacted in the first round by post. This letter contained the URL of the web page containing the questionnaire. The letter was followed up by an email that was sent to two mailing lists. The respondents were asked to answer the questions on behalf of the whole institution and not just for the department they worked in. Further emails were sent two and three weeks later. This time individual messages were sent to those people who had not yet completed the survey.

This exercise resulted in an improved response rate of 18% (114 institutions), three times higher than the response rate in 2003. This is thought to have improved the validity of the conclusions compared with the 2003 survey.

The response to the survey and interest in it was encouraging. The number of respondents who were willing to be contacted for a follow up survey (based on Q43) was high for this type of survey: 41% of those institutions that completed the survey were willing to be contacted again and 78 out of the 103 (76%) who participated were interested in receiving the report.

The fieldwork for the 2006 survey was conducted during February and March 2006.

4.1. Characteristics of the institution and the respondent

4.2. ICT Policy regarding software and its deployment

4.2.1. ICT policy

Two questions were asked in relation to policy; first if the institution had an ICT policy and second if any policy in the institution mentioned software.

As table 3 indicates, 93% of the institutions have a stated ICT policy and 80% have a policy (not necessarily ICT related) that mentions software. Of those institutions that indicate that a policy (ICT or other) mentions software, the majority (80%) say that this is an ICT policy and a minority (20%) say that a different type of policy mentions software.

Table 3. The type of ICT policy/strategy in the institution (Q9 and Q10)

	HE	FE	All
Institution has a stated ICT policy/strategy	88%	95%	93%
Institution has a policy/strategy that mentions software	79%	81%	80%
ICT policy/strategy mentions software	78%a	81%a	80%a
Another policy/strategy mentions software	22%a	19%a	20%a

^a Percentage of those who answered that the institution has a policy that mentions software (answers 1 and 2 in Q10).

Figure 1 shows whether OSS is explicitly considered as an option in institutional policies, and whether its use is encouraged.

Figure 1. Is OSS mentioned as an option when procuring software in policy/strategy? (Q11)

In 2006, for both types of institutions, it is most common that OSS is not mentioned at all (HE=52% and FE=70%). In 2003 there were no HEs that said they had a policy of not using OSS; in 2006 this figure has risen to 4%. Within FEs there was almost no difference between 2003 (3%) and 2006 (2%) in the percentage of institutions that have a policy of not using OSS.

In 2006, OSS is explicitly considered as an option more often in HEs (41%) than in FEs (16%). Some HEs indicated in 2003 that OSS was not only considered as an option but that it was also the preferred option (6%). This is no longer the case (0%), although the reverse happened for FE (0% in 2003, 7% in 2006).

4.2.2. Practice

However, policy might not always be a good reflection of what actually happens in practice, as is shown in figure 2.

Figure 2. Policy and practice of using OSS (Q11 and Q12)

While in 2006 only 25% of policies in HEs and FEs explicitly mention OSS as an option and 63% of the policies do not mention it at all, in practice 77% of the institutions examine OSS as an option. This difference is particularly striking in FEs, where only 16% of policies consider OSS explicitly as an option but 77% examine it in practice.

In general a combination of OSS and proprietary software (PS) is preferred for an institution, as table 4 shows.

Table 4. In the long run, what do you think is better for your institution's computer systems? (Q13)

	HE	FE	All
It's better to use OSS only	0%	0%	0%
It's better to use proprietary software only	15%	13%	14%
It's better to use some OSS and some proprietary	85%	77%	80%
Don't know	0%	10%	6%

Note. Those who did not complete the survey but answered this question are included. Private and other institutions are excluded.

4.3. Skills and awareness of ICT personnel in relation to software

One indicator of the skill that staff have with regard to developing OSS is whether staff submit patches or other code to OSS projects.

Table 5 shows that the number of institutions in which staff submit patches is very small (N=11, 14%) and that often ICT managers do not know whether or not staff submit patches (N=15, 16%).

Table 6 indicates the knowledge (by ICT managers) of the regulations regarding the submission of patches or other code to OSS projects.

For those that do submit patches it is often a working practice rather than part of the person's contract of employment. Again a considerable number of ICT managers (32%) are not aware how the submission of patches is regulated, even where they earlier indicated that they assumed staff do submit patches.

Table 7 indicates that frequently there are only a few members of staff that are aware of OSS.

Staff in HEs have a higher level of awareness than those in FEs. It seems that in general there are a few ‘experts' within an institution, but the majority are not aware of (45%) or only have basic awareness of OSS (35%). The difference between HEs (3.61) and FEs (2.95) in terms of the level of awareness is statistically significant.

Table 8 shows that the same pattern (where HEs score higher than FEs) appears when looking at the level of experience in actually deploying OSS.

In both HEs and FEs there are very few staff with no experience and again, as was the case with awareness, it seems that in most institutions there are a few experts while the rest are unskilled (53%) or only have basic experience (36%).

The responses to Q16 and Q17 show that skill and awareness levels are not evenly distributed between ICT staff members in these institutions (see tables 7 and 8). Although the questions were phrased differently in 2003, there is enough similarity in the responses that the following conclusion can be drawn: while awareness levels were similar in 2003 and 2006, skill or experience levels are higher in 2006.

4.4. Deployment of OSS and Proprietary Software

In 2003 the questions relating to the deployment of OSS were phrased differently, and no distinction was made between the deployment of OSS on servers and desktops. In 2003, 86% of HEs and 72% of FEs deployed OSS. In the 2006 survey the deployment of OSS is without a doubt higher since all institutions indicated that OSS is deployed on desktops and a high percentage indicated that it is used on servers for a number of different applications.

In the 2006 survey, the use of OSS was examined separately for servers and desktops.

4.5. Deployment of software on servers

To understand whether there are intrinsic differences between HEs and FEs in terms of their deployment of OSS on servers, it is important first to understand whether there are differences in the number of servers that these institutions maintain.

Figure 3. What is the approximate number of servers used at your institution? (Q18)

From figure 3 it is clear that HEs have a larger number of servers than FEs. In fact they have almost four times as many. Both in HEs and in FEs the number of servers has gone up from 2003.

Looking back at the difference in student numbers (10952 vs 5430), the difference between HEs and FEs may be accounted by the greater number of students (and thus users) at HEs. Indeed, there is a high correlation (r=.74, p<.01) and probably a strong relationship between the number of students and the number of servers. The relationship between the number of students and the number of servers is stronger in HEs (r=.87) than in FEs (r=.42).

Table 9 shows that most institutions (69%) have deployed, and will continue to deploy, OSS on their servers. There are very few who have tried OSS but do not wish to continue doing so (2%). Some have not used it in the past and will not do so in the future (18%), while another 11% have not used it in the past but intend to do so in the future.

Table 10 shows the future intentions of those who have used only PS on their servers in the past.

Of the 27 (29%) institutions that have not used OSS in the past, 10 (37%) intend to use it in the future. The proportion of HEs who have not previously used OSS but intend to do so in the future is marginally higher than the proportion of FEs (43% compared with 35%).

Those institutions that use OSS on their servers almost invariably use it in combination with proprietary components (see table 11). Only 3% of FEs (2 institutions) say they only use OSS. Even where institutions used OSS in the past but do not intend to do so in the future (see table 10) it is likely that the OSS that was used in the past is still running on these servers and will continue to do so: whilst these institutions do not intend to add more OSS in the future they have not removed the OSS that they have already deployed.

4.5.1. Use of OSS for specific applications on servers

Table 12 shows that, on servers, the use of OSS is most common for database servers, followed by web servers and operating systems. The least common use is for mail software and Virtual Learning Environments (VLEs). In general, PS is used more frequently than OSS with the exception of VLEs in FEs.

Table 12. The use of OSS and PS on servers (Q21 to Q25)

	HE	FE	All
OSS operating systems	60%	54%	56%
PS operating systems	74%**	97%**	89%
OSS mail software	40%**	14%**	23%
PS mail software	100%	100%	100%
OSS web server software	57%	60%	59%
PS web server software	71%	84%	80%
OSS database servers	54%	67%	62%
PS database servers	100%	100%	100%
OSS VLEs	11%**	56%**	40%
PS VLEs	40%	25%	31%

** Difference between HEs and FEs significant at p<.01

These findings are displayed visually in figure 4.

Figure 4. Use of OSS for different server applications (Q21 to Q25)

** Difference between HEs and FEs significant at p<.01

The difference between HEs and FEs is significant with regards to the use of open source mail software, which HEs use more commonly than FEs. In the case of VLE software this is reversed, FEs using OSS VLEs significantly more than HEs. This is particularly important given that HEs are generally more likely to use OSS than FEs. It is one of the few instances in which FEs are more likely to prefer an OSS alternative.

Figure 5 shows that the use of proprietary software on servers is widespread across most applications. As has already been mentioned, it is only in VLEs that OSS can claim an overall lead over proprietary alternatives. Figure 5 also shows that FEs tend to rely more on proprietary operating systems than HEs.

[Page]

Figure 5. Use of PS for different server applications (Q21 to Q25)

** Difference between HEs and FEs significant at p<.01

Table 13 indicates the software that is most often used for each application. It also indicates whether on the whole OSS, PS or OSS and PS is used for this application.

Table 13. Software used on servers for different types of applications (Q21 to Q25)

		HE	FE	All	% only OSS	% OSS and PS	% only PS
Operating systems	Windows Server 2003	74%	94%	87%	0%	56%	32%
	Windows NT or earlier	20%	24%	22%
	Linux	60%	54%	56%
	Solaris	49%**	8%**	22%
	Mac OS X	23%	24%	23%
	Mac OS	9%	10%	9%
Mail software	Exchange	43%	67%	58%	0%	23%	77%
	Novell	17%	14%	15%
	Exim	29%	0%	10%
	Sendmail	20%	10%	13%
Web server	Microsoft IIS	71%	84%	80%	9%	50%	31%
	Apache web server	51%	57%	55%
	Apache Tomcat server	54%	32%	40%
Database servers	Microsoft SQL Server	66%	87%	80%	0%	62%	36%
	MySQL	49%	67%	60%
	PostgreSQL	14%	5%	8%
	Oracle	46%	52%	50%
VLE	Moodle	9%**	56%**	39%	33%	7%	25%
	Blackboard	17%	21%	19%
	WebCT	20%	3%	9%

** Difference between HEs and FEs significant at p<.01

Note. This table lists only the most common responses. The full set of results is listed in appendix C.

For questions Q21 to Q25 there were almost no applications for which only OSS was used; the exceptions were web server software and VLEs. For all other applications both OSS and PS were used by the same institution.

There was a significant difference between HEs and FEs with regard to the operating systems used on servers. Solaris was mentioned significantly more by HEs than FEs. Also, as has already been mentioned, FEs are more likely than HEs to use Microsoft Windows.

In HEs, use of OSS (Sendmail and/or Exim) for mail software (40%) is similar to the usage of Microsoft Exchange (43%), whereas in FEs open source mail software is not used much in comparison to Microsoft Exchange (10% against 67%).

An unexpected finding was that Microsoft IIS is mentioned extensively for web servers (80%), whilst internationally the Apache web server dominates⁶ . Similarly surprising is that Apache Tomcat is used to the same extent (54%) as the Apache web server (51%) in HEs, although in FEs the more common pattern appears, in which the Apache web server (57%) is used more than Tomcat (32%).

There are fewer institutions that use OSS for databases (62% use MySQL and/or PostgreSQL) than use Microsoft SQL Server (80%). This difference is larger for FEs (difference=20%) than for HEs (difference=12%).

The overall difference in the use of OSS between HEs and FEs for VLE applications (see table 12), can be attributed to the Moodle VLE being used far more frequently in FEs than in HEs.

For a number of applications, it was not clear (when designing the survey) which software would be most commonly used. Therefore an open-ended question (see figure 6 and table 14) was used, allowing respondents to indicate which piece(s) of software they used. Most mentioned only one OSS or PS application, and there were no instances in which both OSS and PS were used within the same institution for the same application.

Figure 6. Use of OSS for server applications (Q26)

* Difference between HEs and FEs significant at p<.05

Note. The percentages in the figure indicate the percentage of institutions that uses OSS and not PS for this application

Table 14. What software if any does your institution use in the following areas? (Q26)

		HE	FE	All	% only OSS	% only PS
LDAP	Microsoft	23%	26%	27%	2%	98%
	Novell	29%	15%	20%
	Sun	6%	0%	2%
	Open LDAP	3%	0%	1%
	No answer	37%	52%	47%
Web mail	Outlook	31%	40%	39%	14%	86%
	Novell	17%	12%	14%
	IMP/Horde	11%	1%	5%
	No answer	23%	25%	27%
Calendar/diary server	Microsoft	37%	43%	43%	0%	100%
	Novell	9%	12%	11%
	Oracle	6%	0%	2%
	No answer	34%	29%	33%
Content Management System	Microsoft	9%	7%	8%	11%	89%
	Moodle	0%	6%	4%
	In house	0%	4%	3%
	Other	17%	10%	13%
	No answer	34%	24%	30%
		HE	FE	All	% only OSS	% only PS
Wiki	MediaWiki	6%	6%	6%	100%	0%
	Moodle	0%	6%	4%
	DocuWiki	3%	0%	1%
	None	14%	9%	11%
	No answer	74%	75%	77%
Blog	Moodle	0%	4%	3%	7%	93%
	In house	0%	3%	2%
	Community server	0%	1%	1%
	None	14%	10%	12%
	No answer	77%	75%	78%

Note. These questions did not have closed answers: the respondents were asked to list the components that they used. Percentages are on the basis of the whole sample (n=103).

For this question, all the responses for Wiki software were OSS, whereas only PS were named for calendar/diary server software.

HEs are more likely than FEs to use OSS for webmail (26% vs 5%). For all other types of applications, the differences between the use of OSS in HEs and FEs are not significant. LDAP, blogs and calendar servers are the applications for which OSS is least frequently used.

Table 14 also shows that there is no commonly accepted software for Content Management Systems (CMS). Indeed, the range of software being used for a CMS is wider than for other applications. In all, 69 institutions answered the question about CMS, and they gave 29 different answers. The most used CMS was Microsoft's, but even this was used in only 8% (i.e., 6) of those institutions that answered this question. Whilst FEs mention that they use Moodle as a CMS, it is not mentioned by HEs at all, reflecting the same practice as with VLEs.

Wikis and blogs may be much hyped in the press, but of the 23 institutions (23%) that answered the question about wikis, 48% said their institution did not use them, and of the 22 that answered the question about blogs, 55% indicated that they did not use such software. It can perhaps be assumed that those who gave no answer to these questions also did not use these applications, which would increase these percentages.

4.6. Deployment of software on desktops

A similar set of questions as the ones asked for servers was also asked for the use of software on desktops.

As was the case with servers, HEs have significantly more resources in terms of desktops than FEs (see table 22). Again, this appears to be related to the number of students that institutions have (r=.81, p<.01). As before, this relationship is stronger for HEs (r=.90) than for FEs (r=.82) but the difference between these two types of institutions is less clear than for servers. However, the percentage of desktops that are managed directly by ICT is smaller in HEs (58%) than in FEs (100%).

Figure 7 shows that a large majority of managed desktops run Windows (95%). There is greater use of Windows on desktops in FEs (97%) than in HEs (91%), and greater use of Macs in HEs (14% vs 5%).

Figure 7. What percentage of desktops are managed by ICT services (Q40)

* Difference between HEs and FEs significant at p<.05

** Difference between HEs and FEs significant at p<.01

Table 23 shows that there is more flexibility in HEs than in FEs as, in the latter, the users have less choice regarding the applications that they can run on their desktops.

The differences between the two types of institutions in the way applications on desktops are managed are significant. Looking across the institutions it appears that changing the application set is often not possible (62%). In reality this is the case only within FEs (77%) and not within HEs (30%) where users can change the application set after approval of ICT services (57%).

Institutions were also asked about the applications they deployed on their desktops. The responses are summarised in table 24.

In contrast to servers, the use of OSS on desktops is far less common, both in the past and as predicted for the future. Almost half of the institutions (47%) indicated that they have not used OSS in the past and will not use it in the future. There have been very few bad experiences with OSS, as is evident from the number of responses by those that have used it in the past who are also likely to do so in the future (35% of the total as opposed to the 1% that plans to cease using OSS). Again, there are a fair number of institutions that show a future willingness to deploy OSS without having had the experience of using it in the past (16%). The differences between HEs and FEs in their history of OSS deployment on desktops is not significant.

Figure 8 shows the proportion of institutions deploying only PS, only OSS, and a mixture of PS and OSS on their desktops.

Figure 8. To what extent is OSS used on desktops? (Based on Q33)

Marginally more than half of the institutions (56%) use only PS on their desktops, and it is as uncommon for institutions to use only OSS (2%) on their desktops as it is for them to use only OSS on servers (2%). Amongst HE respondents, there is not one who uses only OSS on its desktops. It is far less common to use both PS and OSS on desktops (42%) than it is on servers (75%). PS thus dominates desktop applications much more than it does server applications.

Appendix A.1 Introduction

Thank you for your interest in this survey. Its main aim is to look at the use of open source and proprietary software in higher and further education institutions across the UK. Although only questions marked with a * are "obligatory", the value of the survey depends upon each respondent completing the full set of questions. So please try to answer each question or sub-question. The survey contains 44 questions and should not take longer than 20 minutes to complete.

This survey asks for factual information about the use of ICT at your institution, and should be filled out by someone who is familiar with ICT across the institution. We will be providing a report on the survey which will be made public. However, any data from the survey will be anonymised in the report.

This survey is conducted by OSS Watch. OSS Watch is funded by the Joint Information Systems Committee (JISC) and is hosted by the University of Oxford as part of its Research Technologies Service. For further information about OSS Watch please visit http://www.oss-watch.ac.uk/ or contact OSS Watch at info@oss-watch.ac.uk.

In these questions, the term 'institution' is used for the whole institution and not just your department. Please provide answers for the whole institution, even if you have to make rough estimates. If you wish to reset a page so that none of the questions have been answered, please use the Reset button that appears at the bottom of the page.

If you would like to receive an electronic copy of the survey results and accompanying analysis, please enter your e-mail address in this box.

Q1 What is the name of your institution?*

Q2 What is the type of your institution?*

Other type of institution (Please specify)

Q3 What is the approximate number of undergraduate students at your institution?*
If you don't know, enter ?? in this box.

Q4 What is the approximate number of post-graduate students at your institution?*
If you don't know, enter ?? in this box.

Q5 What is the approximate number of academic and academic-related staff or teaching staff that your institution employs?*
If you don't know, enter ?? in this box.

Q6 What is the approximate number of other staff that your institution employs?*
If you don't know, enter ?? in this box.

Appendix A.2 Your responsibilities and your technical skills

The following questions ask about your responsibilities within the institution and your technical skills.

Appendix A.3 Management of and policy regarding ICT

What follows are questions about the management of and policy regarding ICT in your institution. Some of the questions refer to open source software (OSS) which is software released under a licence certified by the Open Source Initiative. Typical examples of OSS are Linux (e.g., Red Hat, SuSE, Debian), Apache web server, Mozilla Firefox and OpenOffice.

Q9 Does your institution have a stated ICT policy/strategy?*

Q10 Does this policy or any other policy/strategy at your institution specify anything regarding software?*

Q11 If your answer to Q10 is Yes, is open source software (OSS) mentioned as an option when procuring software in this policy/strategy?

Q12 In practice, is open source software (OSS) examined as an option when procuring software for your institution?*

Q13 In the long run, what do you think is better for your institution's computer systems: to use proprietary software or to use OSS components?*

Q14 Do staff submit patches or other code to OSS projects?*

Q15 If your answer to Q14 is Yes, what are the regulations in relation to staff submitting patches or other code to OSS projects?

Appendix A.4 Awareness and the level of skill in deploying OSS

The questions that follow ask about the awareness and the level of skill in deploying OSS by staff of your institution's ICT services.

Q16 What best describes the awareness of OSS by staff of your institution's ICT services?*

None of these (Please explain)

Q17 What experience of deploying OSS does the staff of your institution's ICT services have?*

None of these (Please explain)

Appendix A.5 Deployment of software on your institution's servers

In what follows you are asked a few questions about the deployment of software on your institution's servers.

Q18 What is the approximate number of servers used at your institution?*
If you don't know, enter ?? in this box.

Q19 Which of the following statements applies best to your institution's deployment of software on its servers?*

Q20 If you are deploying OSS on your institution's servers, to what extent is OSS used?

Q27 If you are using proprietary software on your institution's servers, how important were the following reasons for using proprietary software?

Other reason (Please describe)

Q28 If you are using OSS on your institution's servers, how important were the following reasons for using OSS?

Other reason (Please describe)

Q29 If any decisions were made to exclude OSS on your institution's servers, how important were the following issues in these decisions?

Other issues (Please describe)

Q30 If you are using OSS on your institution's servers, who supports OSS?

Other (Please describe)

Appendix A.6 Deployment of software on your institution's desktops

The following questions are about the use of software on desktops within your institution. By desktops, we mean desktop computers and workstations.

Q31 What is the approximate number of desktops in use in your institution?*
If you don't know, enter ?? in this box.

Q32 Which of the following statements applies best to your institution's deployment of software on its desktops?*

Q33 If you are deploying OSS on your institution's desktops, to what extent is OSS used?

Q35 Which of the following commonly used software are used on desktops and to what extent?
(Please select a button for each line of the table.)*

Q36 If you are using proprietary software on your institution's desktops, how important were the following reasons for using proprietary software?

Other reason (Please describe)

Q37 If you are using OSS on your institution's desktops, how important were the following reasons for using OSS?

Other reason (Please describe)

Q38 If any decisions were made to exclude OSS on your institution's desktops, how important were the following issues in these decisions?

Other issues (Please describe)

Q39 What is the approximate number of desktops that are managed by ICT services?*
If you don't know, enter ?? in this box.

Q41 Which of these statements is most appropriate for desktops managed by ICT services?

Appendix A.7 Other information

The following questions obtain the answers to some final questions.

Q42 Is there anything you would like to add to the information that you gave in this survey and that you have not been able to express?

Q43 We might want to follow up on this research project. Would you be willing to be contacted for participation on further research into this subject?*

Appendix D.1 Contacting respondents

The method used for contacting ICT managers in this survey was very successful and it is highly likely that with more time for the fieldwork the response rate would have been even higher. Spam emails seemed to have been very efficient and showed an immediate increase in responses.

Appendix D.2 Measuring institution size

Table 1 gives the average number of students and staff (produced from the answers to questions Q3, Q4, Q5 and Q6). There was a problem with the responses to these questions. The number of institutions that replied to these questions is shown in Table 34.

It is clear that the postgraduate question is problematic for the FE institutions, but there also seemed to be a problem for those in HEs; of the 30 institutions only 20 answered this question. The questions in relation to staff were less problematic and most missing answers were caused by the person not knowing this number. It has to be noted that there was great variance in the number of students at both types of institutions.

In future research, it might be useful to look at student-staff ratios instead of the size of the institution.

Table 35 shows that the student/staff ratio is lower in HEs than in FEs, but also that there is a lot more variety between the FEs than between the HEs (based on the standard deviation). However, the difference in student-staff ratio is not significant. In future research with a hopefully higher number of institutions and a better formulation of the question about postgraduate students this indicator might be more useful.

Appendix D.3 Question wording recommendations

In a possible follow-up survey the questions that were left open ended in terms of applications on servers can be categorised according to the categories found in this survey, with the exception of the question on Content Management Systems.