Whitfield School August 2006 Whitfield School 1: 1 Computing Program: An assessment of the implementation of an open-source 1: 1 computing model Table of Contents

Unique Considerations Associated with Open-Source 1:1 Computing Programs

Download 152.27 Kb.

Page	3/4
Date	31.01.2017
Size	152.27 Kb.
	#13454

1 2 3 4

Unique Considerations Associated with Open-Source 1:1 Computing Programs

While the Whitfield implementation had many of the same themes seen in the previous 1:1 Computing program research, the assessment did uncover several somewhat unique considerations associated with the Whitfield open-source environment. We have organized our discussion of these unique considerations into four sections as follows:

Total cost of ownership (TCO);
Program adoption;
Program impacts; and
Technical problem resolution.

Total Cost of Ownership

Even before initiating its migration to the 1:1 computing model, Whitfield already had consciously decided to maintain and support a relatively technology-intensive environment. To illustrate, during 2004-5 the student/client computer ratio already was down to 1.92 for the school as a whole and 1.89 for grades 11-12. Teachers already had their own computers (primarily laptops). The IT support function consisted of three full-time staff, plus the equivalent of a full-time position that was outsourced. (This position was brought in-house in 2005-6.) Compared to many other schools (especially public schools) this is a relatively high number of support positions for a total end-user population of approximately 550 (about 460 students and 90 faculty/staff).

Whitfield’s level of technology investment was a natural development given its educational model that focuses on providing an innovative curriculum and educational environment in which teachers and students are not afraid to experiment, and teachers are facilitators and partners in learning with students. In this model, the focus is on helping students learn how to think, research and analyze information, and develop into lifelong learners. Whitfield also aims to prepare students to function effectively both in the higher education and work environment of the 21^st century, and application of innovative uses of technology is viewed as an important contributor to achievement of this goal. In fact, one of Whitfield’s specific educational goals is to “Seek opportunities to use technology to access diverse resources and communities as a means to participate in a wider community of learners.” By placing as much computing power and flexibility in the hands of students as possible, Whitfield decided that it would be much more likely to achieve these ultimate objectives for its students.
When Whitfield’s management team decided to implement the 1:1 computing model, their desire was to maintain their programmatic quality and flexibility, and the high level of IT support, in as cost-effective a manner as possible. The decision was made not to attempt to fund the movement to this new model through an increase in tuition or a technology fee assessment on the parents of the students who would be receiving the laptops. In part, this was to maintain Whitfield’s overall cost competitiveness versus other local area private schools. More importantly, however, it wished to convey the message to parents that providing personal computing resources for each student will, from now on, be as integral a part of their education process as any other expenditure item that is covered by the tuition payment.
Given this challenge from the top management team, Whitfield’s IT management proposed the combined Citrix/Linux approach, both for economic and pedagogical reasons. The economic reasons included the following:

Reduced risk and IT support staff time due to the ability to concentrate management of the Windows environment on centralized servers;
Reduction in software license fees due to pooled usage of Windows applications through use of the Citrix environment;
Acquisition of laptops with a less expensive configuration, and a longer expected life (four versus three years) due to avoiding the need to load and maintain Windows applications;
Ability to economically obtain significant Linux-based applications at a relatively low incremental software cost; and
Overall IT support staff time and cost savings due to the relatively low security risk and operational stability of the Linux environment.

The use of Linux also presented significant advantages from a pedagogical perspective. It provided the flexibility for teachers and students to try out and add new products, generally at minimal cost, without building a dependency on a particular proprietary model (i.e., Windows). It also better reflected the learning models that are commonly in use in higher education organizations – many of whom have adopted and emphasize the use of Linux and open source applications in their own computing environments.

In order to manage Whitfield’s 1:1 computing assets as carefully as its other IT equipment investments, Whitfield decided that laptops would remain school property, to be returned by students at the end of every year for maintenance and product upgrades and then recycled for use by another student during the next school year. Whitfield also committed to a regular refresh program, thereby indicating its long-term commitment to the program. The decision also was made to invest in a relatively durable laptop computer with a high-quality, permanent carrying case to guard against potential student misuse and reduce the frequency of repair. This was one desirable feature of the Lenovo ThinkPad model that Whitfield selected for this initiative.
Table 5 summarizes Whitfield’s overall technology costs per client computer for both 2004-5 and 2005-6, and compares these with the case study ranges provided in the CoSN/Gartner model.
Table 5: Comparative Total Costs per Client Computer: 2004-5 and 2005-6

Costs per Client Computer	2004-5 School Year		2005-6 School Year		Case Study Range
Costs per Client Computer	Grade 11 - 12	All Grades	Grade 11 - 12	All Grades	Low	High
Direct Costs	$ 1,481.86	$1,491.42	$ 1,394.72	$ 1,722.91	$384.72	$1,241.86
Indirect Costs	$ 92.90	$ 92.90	$ 62.15	$ 56.39	$131.00	$2,012.93
Total Costs	$1,574.76	$ 1,584.32	$1,456.87	$ 1,779.30	$1,004.00	$3,254.79

One of the indications from this table is that, even before introduction of 1:1 laptop computing, Whitfield is characterized by a relatively high level of direct costs, but a low level of indirect costs, on a per-client computer basis when compared with the CoSN/Gartner survey sample. This is logical given the emphasis that the school places on providing high quality, appropriately-staffed IT support services to its students and teachers. The low indirect cost figures indicate that there is a payoff in that end users at Whitfield apparently expend comparatively less time than their peers elsewhere in maintaining and troubleshooting their own computing equipment. Presumably, this provides them with the opportunity to devote relatively more of their available time to higher value-added teaching and learning activities. On an overall total cost basis, in fact, the average per-client computing cost at Whitfield appears to lie nearer to the low end than the high end of case study range.

The data in Table 5 also indicate that, as 1:1 computing was introduced in Grades 11 – 12, the cost per client computer moved from being near the schoolwide average in 2004-5 to over $300 below the average in 2005-6. In fact, viewed from a total cost perspective, Whitfield achieved a net increase of 119 student laptop computers in Grades 11 – 12 between 2004-5 and 2005-6 at a total incremental cost of $115,943, or $974 per computer. While there were other factors that influenced these cost movements (e.g., equipment upgrades in other grade levels, transfer of some desktop computers and their associated costs from grades 11 – 12 into lower grades, and certain non-recurring administrative systems costs in 2005-6 that needed to be allocated across the entire end-user computing base), there is a clear indication that the 1:1 computing model as introduced at Whitfield achieved management’s expectations for cost efficiency.

Table 6 provides a breakdown of Whitfield’s direct computer costs on a per-client computer basis by major category (hardware, software, direct labor and external application providers) for 2004-5 and 2005-6.
Table 6: Comparative Direct Costs per Client Computer: 2004-5 and 2005-6

Costs per Client Computer	2004-5 School Year		2005-6 School Year		Case Study Range
Costs per Client Computer	Grade 11 - 12	All Grades	Grade 11 - 12	All Grades	Low	High
Hardware	$519.79	$519.29	$573.76	$605.98	$176.65	$432.12
Software	$177.61	$177.44	$189.18	$268.31	$6.00	$218.76
Direct Labor	$784.46	$794.69	$631.78	$848.62	$180.60	$818.57
External Application Providers	$0.00	$0.00	$0.00	$0.00	$7.08	$93.00
Total Costs	$1,481.86	$ 1,491.42	$1,394.72	$ 1,722.91	$ 384.72	$1,241.88

One of the first conclusions from viewing this table is that Whitfield’s costs on per-client computer basis are relatively high in comparison to the published CoSN/Gartner case study ranges. This is not surprising, nor is it of particular concern in itself to Whitfield’s senior and IT management team. Compared to most other schools or districts (particularly public schools), Whitfield maintains its computing equipment in current condition, operating under a three-year refresh cycle for desktop computing and now a four-year cycle for laptops. Schools or districts that do not have such a plan will tend to have older computers whose costs have been fully amortized and that therefore do not appear in this model.^² In addition, as stated earlier Whitfield by choice maintains a well-staffed and knowledgeable IT support group that results in relatively high costs but also high service levels when compared with its peer institutions.
Of great interest to Whitfield management is whether or not the move to 1:1 computing held the line on additional costs so that the school could better absorb the impact of this addition to its technology base within its current financial structure, and without the need to increase tuition or charge an additional fee. From this perspective, Table 6 provides evidence that this management objective was substantially achieved. During 2005-6, Whitfield experienced some one-time administrative computing costs that effectively amounted to a surcharge of $126.30 on a per-client computer basis. This helped contribute to an increased cost per computer for the school as a whole between 2004-5 and 2005-6. Despite this, the overall per-computer cost for Grades 11 – 12 actually declined between the two years. With further analysis of the detailed cost components as reported by the model, the major contributors to the overall decline appear to be as follows:

Decline in server and network costs of approximately $85 per client computer, primarily due to the ability to leverage the Citrix/Linux model to support and manage a larger end-user computing base;
Decline in software costs of more than $100 per client computer, when adjusted to remove the effect of the one-time administrative computing costs. This reflected the movement to centralized management and sharing of Windows software licenses with the Citrix computing model, as well as the low costs of obtaining software in the Linux operating environment; and
Decline in direct labor costs of more than $150 per client computer, mostly concentrated in IT operations staff savings due to their ability to manage a large user base. The TCO results indicate a client/IT staff ratio of 122.3 in 2005-6 for Grades 11-12 versus a ratio of 90.5 for the school as a whole in the same school year, and a ratio of 83.0 for Grades 11 – 12 in 2004-5, prior to introducing the 1:1 model. This reflects the relative stability and security from outside intrusion of the managed Citrix environment and the Linux operating system, the high availability of Citrix (99.996% was achieved during the year), and the reliability of the laptop computers and their configurations. Regarding the last point, Whitfield’s IT service records indicate that 13% of its laptops needed to be reimaged during the year due to hardware or software failure, versus 40% reported by other peer schools. Moreover, Whitfield recorded insurance claims on only 5% of its laptops versus 28% at peer schools.

On the other hand, Whitfield did experience an offsetting increase in hardware costs per client computer of $136 as it transitioned grades 11 and 12 to the 1:1 computing model. This appears to be the result of replacement of older, already mostly amortized, desktops with newer laptops that also tend to carry a slightly higher price tag for equivalent functionality. Whitfield also paid for some additional investments, such as a carrying case, to help protect these new assets and extend their useful lives. Therefore, the first-year of the 1:1 computing experience at Whitfield did not result in a direct hardware cost saving at the client computing level.
Table 7 provides a same-year (2005-6) comparison of per-client computer costs between the grade levels at Whitfield (11-12) that did use the 1:1 model and the grade levels that did not (6-10).
Table 7: Comparative Direct Costs per Client Computer: 1:1 versus Traditional Computing Model (2005-6)

Costs per Client Computer	2005-6 School Year		Case Study Range
Costs per Client Computer	Grades 11 – 12 (1:1 Model)	Grades 6 – 10 (Traditional Model)	Low	High
Hardware	$573.76	$637.37	$176.65	$432.12
Software	$189.18	$345.42	$6.00	$218.76
Direct Labor	$631.78	$1,059.99	$180.60	$818.57
External Application Providers	$0.00	$0.00	$7.08	$93.00
Total Costs	$1,394.72	$2,042.78	$ 384.72	$1,241.88

This table indicates a striking difference in terms of higher per-user costs for the traditional model in all major categories. The software costs are especially different due to the need to maintain a full set of software licenses on each workstation that supports Grades 6 – 10 (these also operate in Windows, not Linux). The higher direct labor costs for the traditional model reflect, in part, the frequency of repair and troubleshooting required by IT staff for the desktop computers that reside in a lab environment. In addition, to the extent that IT staff time requirements are a function of the number of users instead of the number of machines supported, the student/computer ratio of 2.45 in the traditional model is a contributing factor.

Whitfield’s IT management team plans to continue the rollout of the Linux/Citrix 1:1 computing model to its remaining five grade levels over the next few years. Through their own analysis of their 2005-6 acquisition costs, and comparison to the probable costs of implementing a fully Windows-based 1:1 model, the team estimates that they saved nearly $30,000 in costs, primarily due to being able to purchase laptops with a leaner configuration and also due to insurance savings.^³ In future years, IT management plans to achieve additional per-computer savings by increasing the computer/software license sharing ratio in the Citrix environment for Windows applications. Upon the completion of the full rollout, IT management’s projection is that annual savings will approximate $200,000. Whitfield’s senior and IT management also are committed to ongoing identification and realization of additional economies, so that characteristics of the Citrix/Linux 1:1 computing model can be anticipated to evolve significantly over time.
Program Adoption

Prior to Whitfield, 1:1 Computing Programs have employed standard Windows-based operating systems, which offer a familiar interface for students and faculty alike. Whitfield, on the other hand, leveraged the open-source operating system, Linux, along with open-source software products. Anticipating that some users would be more comfortable with Windows and Windows-based applications, the school included Citrix on all the laptops as an additional, or alternative, option.

Since Windows continues to be the most prevalent operating system on home and business computers throughout the world, it stands to reason it would tend to promote a fairly seamless transition for schools. With most users already familiar with its design and functionalities, a Windows-based implementation comes with a certain degree of automatic comfort and acceptance.
In contrast, open-source technologies still represent a small (but growing) fraction of the computing community. While they seem to be on the rise and are beginning to close the gap with their competitors, their reach is not widespread among the general public. Therefore, it could be expected that Whitfield students and faculty would have less initial familiarity with open-source computing—potentially increasing the effort required of a student or teacher to adopt this program.

As the baseline data illustrates, students at Whitfield did have a lower degree of initial comfort with open-source features, such as the Linux command line. At the outset, only 42% of students indicated they were very comfortable or comfortable with it, while 15% said they were not sure what the command line was. However, by the end of the academic year, students reported a significant difference in their comfort level with the Linux command line. In Phase II, 60% of students reported being very comfortable or comfortable with the Linux command line, and only 6% were still not sure what it was (p<.05). See Table 8 for the results.

Table 8: Comfort using the Linux command line

One could also argue that the initial 42% of students comfortable with the Linux command line actually represents a fairly high starting point. This finding may indicate that students today are relatively technology savvy and may already be knowledge about open-source technologies. Further evidence of this comfort with open-source technology can be found in the time students spend using it. Here, the Whitfield experience is instructive, because students had the option to use Linux or Citrix. In our surveys, students indicated they spent more than half their time working with Linux (Phase I=59%, Phase II=52%). This finding provides further evidence that the Whitfield model received fairly broad adoption among the student population. See Table 9 for the results.

Table 9: Use of Linux versus Citrix

Program Impacts

One of the well-documented impacts of 1:1 Computing Programs is the changing educational roles it fosters, whereby students act as teachers, and teachers act as learners and facilitators. Part of this adjustment clearly is technology-related as all the parties work to become proficient with their computers and software applications. Whitfield’s implementation model accentuates this need to change, given the greater initial lack of familiarity with open-source technologies. Nevertheless, students and teachers will, over the course of the year, find ways to meet this challenge and acquire the skills necessary to operate effectively with their new laptops. Early adopters or skilled users of these technologies become the ‘go-to’ people for help, and in this way, teach others collaboratively. One faculty member stated, “[The students] know who the experts are in class, as far as the technology goes, and they’ll go to them.”

In a related way, Whitfield’s model helps its students and faculty improve their technology competence—both in terms of general computer ability as well as skills with open-source technology. With open-source operating systems and software on the rise – and showing strong promise – these experiences could prove extremely valuable now and in years to come. One faculty member stated, “I didn’t learn enough about Linux, but we’re so rooted in it now that it would be hard to go back…but I’m willing to become more versatile.”

As expected, many of the changes in teaching associated with Windows-based 1:1 Computing Programs were observed at Whitfield. However, the faculty reported that some types of lessons were limited by the available open-source software. For example, the open-source analogue for creating movies was described as “not in any way, shape or form” comparable to the commercial product. These issues are probably partly a function of the novelty of the open-source product, and they may be mitigated over time as the software quality improves.

Technical Problem Resolution

The 1:1 Computing Program at Whitfield provides interesting evidence of the impact open-source technologies may have on the incidence and resolution of technology-related issues. Initially, one might expect significantly more technical problems with Linux, since it would likely be less familiar to students and faculty members (as described above). However, students reported only slightly more technology-related issues with Linux when considered in proportion to its use versus Citrix (or Windows). Further, all technical issues with Linux were related to personal laptops, while most of the issues related to Citrix occurred on the school’s highly controlled servers.

In addition, the technical issues associated with Linux appear to be more easily resolved by the student himself/herself, by a peer, or by a teacher. Only 11% of issues with Linux needed to be addressed by the Information Technology Department, whereas with Citrix, 18% of them required IT intervention. However, there are a higher percentage of issues that remain unresolved with Linux than Citrix (8% vs. 3%). Overall, these findings seem to suggest that Linux may generate slightly more technical issues, but they tend to be resolved with less need for IT involvement. See Table 10 for the results.
Table10: Technical problems: Linux versus Citrix

Some of these technical issues may have been due to the assumption that students would be spending more time using Windows applications. As such, the need for more extensive open-source training was somewhat underestimated. Therefore some of these technical issues also may be mitigated by a more thorough up-front training for students and faculty. For various reasons, Whitfield faculty members were unhappy with much of the initial training on the new 1:1 Computing Program. Comments such as “The delay in getting the machines imaged…everything got crunched down to a few days before school started and we had to turn around and help our kids use it, which was completely unreasonable” and “We hadn’t had any chance to play with Linux; we maybe had 2 to 3 days...” and “We recognize there has not been enough training” were abundant.
This lack of necessary training may have partially contributed to the volume of issues encountered in conjunction with the introduction of Linux-based laptop computing. Therefore, it appears up-front training on Linux and open-source software products – perhaps even more than would be required for a Windows-based approach – would make their adoption more seamless and perhaps result in fewer issues.

Download 152.27 Kb.

Share with your friends:

1 2 3 4