Why has the Peru One Laptop Per Child program failed to live up to its promise?

The full paper is reproduced below excluding the Bibliography which is found via the PDF download

HD PDF New Why has the Peru One Laptop Per Child program failed to live up to its promise? (3018)

Author: Matt Haikin  

1.  Introduction

It is difficult to evaluate effectively ICT-Education projects due to the long timescales and longitudinal studies needed to assess impact, and the difficulty in linking cause to effect. This is especially relevant with innovative programs, such as One Laptop Per Child (OLPC), that offer a powerful and seductive aim of significant transformation.

Results from OLPC programs have been mixed, compounded by problems with delivery and lack of robust evaluations, making it challenging to assess whether there has in fact been any transformational impact at all.

This paper uses the Design Reality Gap framework to look at the Peruvian rollout in more detail, demonstrating how delivery problems have obscured much of the potential of its innovative underlying approach. It concludes that radical changes are necessary for future phases of the project in order that the potential social impact of OLPC can be properly assessed and evaluated.

2.  The One Laptop per Child programme (OLPC)

2.1 What is One Laptop Per Child?

Recently there has been a growing interest in “1-1 computing” initiatives –most well-known are those of the One Laptop Per Child Foundation (OLPC).

The One Laptop Per Child Foundation is a US-based charity established by Nicholas Negroponte in 2005 to “empower the world’s poorest through education”(OLPC, 2011a). OLPC set out to achieve this by designing a laptop computer specifically for the conditions in developing countries, and partnering with the governments in these countries to rollout the laptops to all their schools and children.

The foundation and Negroponte have a view of learning known as constructionist learning – the theory that individual learners/children construct their own mental models to understand the world around them. The OLPC hypothesis is that they can do this most effectively with a personal learning device (OLPC laptop) with them at all times, hence the 1-1 model where each child owns their own laptop in perpetuity. Implicit in this approach to learning is a learner-centric, facilitative approach to learning, rather than a more traditional instructional approach.

Over 2,000,000 XO children and teachers in Latin America are part of an OLPC project, with another 500,000 in Africa and the rest of the world (OLPC, 2011b). Although these figures are for planned deployment not necessarily laptops in current use, it is a major program and it is important to understand its successes and its impact.

2.2 How should OLPC meet educational development goals?

There are two of UNESCO’s six Education for All (EFA) goals that one would expect the OLPC programmes to have a significant impact on:
Promote learning and life skills for young people and adults
Improve the quality of education

This is because the key claim of the programme is that it will motivate young learners and provide them with the “opportunity and resources” to unlock their potential” (OLPC, 2011a) through integration of technology into the education process.

In Peru’s case these goals were translated into the following specific program goals:
• Improve quality of primary education, especially in poor rural areas
• Integrate XO in pedagogical practice to achieve national curriculum abilities
• Train teachers in pedagogical use

3.  The OLPC implementation in Peru

Peru is a big country with a difficult topography for broadband, with serious inequality problems, with indigenous communities speaking different languages, often in isolated rural areas, making the Peruvian rollout especially challenging. Peru also chose to focus its rollout on isolated, rural areas (with high proportions of indigenous language speakers), many with ‘single-grade schools’ with only one teacher – who may never have used a computer before (Breitkopf, 2011) – with virtually no support staff or other infrastructure available; adding to the challenge it faced.

The funding for the rollout in Peru was healthy, but this was focused primarily on the costs of the hardware and deployment, with no budget for creating content, improving connectivity, and very little for planning/admin.

This is evident by comparing to the Uruguay costs below.

Peru’s rollout model is unusual and focuses around the idea of ‘local autonomy’ (supporting the view of the World Development Report 1988 that “governments should decentralise education to give more power to those with the most information about educational needs and how to meet them: students, parents, teachers and local school administrators”.

It engaged with teachers and expecting teachers and students to work out the best way of using the laptops in their own local contexts (Severin & Capota, 2011). This is one of the most interesting factors in the Peruvian rollout – by following best practice on the role of teachers, and in stark contrast to the model in Uruguay, where the top-down approach was criticised as a marketing exercise that bypassed educators (Daina Beitler 2011, LSE, email, December 14th) it had enormous potential for success.

The following sections will explore why the majority of this potential success has not been realised.

4.  Success/Failure and impact on development goals

So far there is very sparse evidence to show whether any positive impact is taking place against any of these educational development goals, due to “short time spans, lack of appropriate evaluation methodologies, and lack of commitment to study impact” (Severin & Capota, 2011). However, some data is available:

An early evaluation in 2009 reported a 50% improvement in reading comprehension and almost 60% improvement in textual and mathematical analysis, leading to a shared UNESCO prize for the ‘Use of Information Technologies and Communication in Education” award (OLPC, 2011c). However, most reports since then are vague and contradictory.

There are reports of attendance and motivation increasing, of more positive attitudes, of over 90% of both teachers and parents saying the laptops have improved the quality of learning (Santiago, 2010). However, there are also reports that, in many cases, use of the laptops declined after about 3 months. There are no statistics or thorough evaluations available yet that tie back specifically to the OLPC goals in Peru, or more widely to the EFA goals.

This may not be the real problem however. It is always difficult to assess short-term impact, and even more difficult to tie this in any causal way to any long-term transformational impact (where ICT projects hopes lie), and OLPC is a relatively new program so the lack of quality evaluations at this point in time is to be expected.

More worryingly, assessing any impact has been made significantly more difficult due to avoidable delivery problems, which will be explored in the following section.

5.  Design Reality Gap analysis of implementation/delivery

Traditionally a design reality gap approach to project success looks at the difference between ‘where we are now’ and ‘where the project wants to get us’ (i.e. assumptions built into the project design) across a range of dimensions (most commonly the ITPOSMO range : Information, Technology, Processes, Objectives/values, Staffing/skills, Management systems/structures, (other) Resources) , asserting that the larger the gap (in any dimension and overall) the more prone to failure a project is:

One of the criticisms levelled at a DRG approach that is especially relevant to this paper is how subjective it is (Heeks, 2008) – this has been mitigated somewhat by drawing on a wide range of sources to define the success factors, but the judgement of the gap is, by necessity, personal. One way to mitigate this further (outside the scope of this paper) might be to interview various stakeholders of the projects and get them to assess the quality, nature and causes of the gap themselves – while still subjective, the quantity and nature of the participants would give more relevant and better quality results.

5.1 Using a Design Reality Gap approach in this context

The DRG approach has been adapted slightly here, as the OLPC projects have been going for a number of years; the designs and plans have changed during this time, and are continually evolving. Therefore, while the design-reality gaps in each dimension at the start of the programme are interesting, more relevant is the size of the gaps still remaining after the project has adapted, and the success/failure in closing the initial gaps as the programme has evolved.

5.2 Customising the DRG framework for this purpose

While this generic approach to Design Reality Gap analysis gives a good starting point of which areas to focus on when analysing a project, it is possible to customise the dimensions to better suit the project being analysed.

It is widely accepted that the worst thing to do in ICT and Education is “Dump hardware in schools, hope for magic to happen” (Trucano, 2011), so what is needed for this DRG framework is an outline of the other critical factors that need to go along with the hardware.

Drawing on a range of literature on success factors in ICT4D and eGovernment projects in general (Heeks, 2002, 2003; Heeks & Walton, 2011), in ICT & Education projects more specifically (Brunello, 2010; Hosman & Cvetanoska, 2010; Tinio, 2003; Warschauer, 2003), and also success factors learned from 1-1 laptop initiatives globally over the last few years (Nugroho & Lonsdale, 2010; Severin & Capota, 2011), the following customised version of the ITPOSMO framework has been developed:

While there are clearly other important factors within each of these dimensions (including of course the XO laptop itself), the above seem most relevant to an analysis of the OLPC programme in Peru.

Two other critical issues came out in the majority of the literature: Evaluation and Sustainability. These fall outside of the DRG analysis and will be returned to at the end of this paper.

5.3 Analysing Peru’s OLPC programme in a DRG framework

(Note: gap score of 0 = no gap, 10 = huge gap)

5.4 Initial Gaps and Reduced Gaps

We can see from these diagrams that OLPC Peru started with significant gaps in at least five dimensions. This gave it an extremely high chance of failure.

To its credit, the project has since managed to reduce these gaps significantly (In some cases this by changing the designs, e.g. providing more digital resources and curriculum; in some cases changing the ‘reality’, e.g. by providing better training to up- skill teachers). This has greatly increased the chances of success but there are still medium gaps in at least four dimensions which is still an unacceptably high level of risk.

In particular, there is still a moderately high gap in the Technology/Infrastructure dimension, which is such a critical part of the rollout that it continues to risk jeopardising any potential positive impacts going forward.

These high design-reality gaps have added to confusion assessing the failure to demonstrate significant positive impacts. This is partly due to lack of evaluation, also to mismatched expectations of ICT4D and ICTE projects (a failure of the constructionist approach proposed by Negromonte and the OLPC foundation). However, the analysis above demonstrates that even with the best evaluations in place, these implementation problems in the rollout would obscure the potential impact of the program; failures are as likely to be due to, for example, connectivity problems, as to an underlying problem with the OLPC approach.

Returning to the social and developmental impact – from the above analysis we can understand better the reasons behind the lack of impact identified at the start:

Integrate XO in pedagogical practice to achieve national curriculum abilities; and Train teachers in pedagogical use
It is immediately apparent that, despite these being core goals of the programme in Peru, this training was simply not provided, so it is hardly a surprise that these goals failed and teachers do not utilise the laptops to their full potential

Improve quality of primary education; especially poor rural areas
The modest and contradictory results can certainly be tied, at least partially, to the implementation problems, especially the content, connectivity and technical issues meaning in rural areas many of the laptops were not utilised effectively

Given the vast sums of money being spent on OLPC, and the exciting potential for transformation it espouses, this is extremely disappointing. The following section goes on to look at how Peru could learn from and improve on this so that future evaluations can more accurately assess the social impact of the OLPC approach.

6.  Recommendations for best practice

Significant challenges remain for the future of Peru’s OLPC programme. There are still high design-reality gaps remaining relating to Technology (infrastructure and connectivity), Processes (technical support) and skills (teacher training), with a potentially large gap around community involvement, although little clear information is available on this dimension. Some key recommendations are outlined below.

Get the nuts and bolts right
It is clear from ICT4D literature that there is already a well understood and accepted set of success factors relating to delivery and implementation of this type of project. Governments should be aware of this, and ensure rollout plans take account of every factor, with adequate resourcing, time and skills. Critical factors in the case of Peru are below, compared to best practice from Uruguay which is regarded as a best practice example of delivery (although with significant problems in other areas such as pedagogy and participation):

• Connectivity
The lack of internet access, and the inadequacy of USB-sticks as a replacement compounded a lack of content to turn the laptops into little more than brief novelties for many teachers and students.

In contrast, Uruguay rolled out laptops to places which already had good connectivity, and partnered with ANTEL (state-owned telecoms company) to offer subsidised wireless access in public spaces (Derndorfer, 2010b).

While Peru’s terrain may mean it is not possible for to rollout connectivity to all isolated rural areas, but these types of partnership could help accelerate the process, and a more robust model based around USB-sticks could also help.

• Processes
The support and repair infrastructure was totally inadequate and only a partial solution has been implemented. A model used in Uruguay involved a nationwide partnership with local, repair shops (trained to service XO laptops), and mobile repair teams able to repair many problems on-site (Derndorfer, 2010b). The size of Peru makes this more challenging, but an adapted version of this solution could help.

• Community Involvement
Many of Peru’s problems with inadequate learning content could have been resolved by working with external partners. In Uruguay, Plan Ceibal has been extremely receptive to partnering with community-led initiatives, with at least three external community-led projects – a University offering community outreach, an NGO developing e-learning materials, and a network of volunteers providing much needed support). As well as providing an opportunity to create more e-learning content, the density of NGOs in Peru means could provide an invaluable way to support more remote communities.

• Teacher Training
Although Peru may be at less of a disadvantage than many countries, due to the similarity in its constructionist pedagogy and that of OLPC, this is still a major shift in the approach to teaching, and the level of training was clearly inadequate in all areas – from the use of ICT to its integration into the curriculum and pedagogical support.

This seems shared across OLPC rollouts. Uruguay re-designed its training program at least three times, there is still feedback that teachers do not feel engaged and feel they need more support in integrating the technology into curriculum (Derndorfer, 2010b). It seems that these programmes consistently under-estimate the time, buy-in, work and support/training needed to get teachers to be confident in use of technology and associated new pedagogy of “mentor, supervisor, facilitator for peer learning” rather than the instructor-led paradigm dominant in most of the world (Brunello, 2010).

But pedagogy change is slow. Teachers already play different roles in their work – instructor, demonstrator, manager, consultant, questioner, observer, co-learner (Loveless, DeVoogd, & Bohlin, 2002) but need support to enable them to use these skills in a new context. If they are to shift management and control of activities to children as the new pedagogy requires, they must be treated sensitively, participate in policy and planning, delivery and assessment, and be given adequate training and support (Loveless, 1995).

A Bigger Role for Government
The World Bank suggest an active and collaborative role for government in relation to changing the way educators work with technology – adapting curricula, disseminating research on how children learn, exposing teachers to new technologies and providing in-service training opportunities (World Bank, 1998). In stark contrast to this, the findings above demonstrate that Ministry of Education in Peru relied on the existing infrastructure, and changed very little else.

‘Sustainability failure’ is something that seems to particularly affect developing countries (Heeks, 2002); perhaps more-so in the current economic climate. The justification of a 1-1 computing program hinges around access for all. If this is not practical and sustainable, the government should ask if it is sensible to spend so much time and money on something that will not be around for long enough to achieve long-term transformational aims, when the same money could have been spent on something already known to work.

As outlined at the beginning of this paper, without robust evaluation of social impact, nobody will know what the benefits of this new approach to technology in education truly are. Even if, at present, the delivery issues above would obscure these findings.

Fixing the delivery problems is just the first step – after this is done, evaluations become increasingly important as only at that point can they begin to draw conclusions around the true purpose of the programme – whether this type of 1-1 laptop initiative, combined with a constructionist approach to learning, can produce significantly better results than a more traditional approach to education, and is worth the vast sums of money being spent on it.

7.  Conclusion

The IAB report questions whether “participation in a 1-1 program improves student employability, income, career paths and performance are still unanswered” – clearly further evaluation and longitudinal study is required, but this paper has demonstrated that – unless and until we get the delivery right, the results of these evaluations will continue to remain inconclusive.

On a more positive note, we know what to do to improve this situation – and in fact always have done – solve the delivery and training problems, involve the teaching profession from day one, evaluate continually, and ensure the program is sustainable.

Once this is done, we can begin to understand whether the new technology-driven paradigm of 1-1 learning is a powerful tool to help the developing world and poor communities, or a chimera wasting enormous sums of money that could be put to better use elsewhere.

HD PDF New Why has the Peru One Laptop Per Child program failed to live up to its promise? (3018)

Tags: , , ,

Leave a Reply

This blog is kept spam free by WP-SpamFree.