Back in May we published an article called "How to learn when 53 percent of XO laptops are broken?" which led to some interesting discussions as well as some people becoming quite unhappy with olpcnews. The core of said story was that a report from Uruguay indicated that more than half the XOs in the school that had been sampled were in an unusable state. I think it's needless to say that such a high number of failures - if it turned out to be true on a big scale that is - would be quite scary.

Broken keyboard on Uruguayan XO

Hence one of the objectives for my time in Uruguay was to get a better idea of just how many XOs really are broken. Needless to say that I don't really have any new and conclusive numbers to share but after talking to several people, from inside as well as outside Plan Ceibal, I think I at least have somewhat of an understanding of what's going on.

The numbers

According to an official report which received quite a bit of media attention last week 14,2% of the XOs in 275 surveyed classes at 55 different schools (out of more than 2000 schools participating in Plan Ceibal) where unusable due to some form of breakage. An additional 6,2% of XOs was currently undergoing repairs and hence unavailable for the children. Together with machines blocked by the security system or a full Journal (3,9%) and ones where the state was either unknown or something else (3,1%) this means that 27,4% of the XOs weren't in a usable state.

What the numbers mean

Given that the report states that it worked with a representative sample of schools both in as well outside of Montevideo I'd say that the numbers are likely to be a pretty accurate picture of the state of XO breakage in April 2010. However admittedly without further data about the surveyed schools it's hard to say just how representative the sample really is. (Please see the update at the end of the article for more information.)

Another interesting bit of information that I almost missed is that the XOs in the interior of the country were in most cases handed out in 2008 whereas the schools in Montevideo received their XOs throughout 2009. While not being the single cause of the lower number of functioning laptops outside of Montevideo I would assume that the extra year of usage is one of the core factors here.

The report also points out that a number of measures have since been take to deal with the issues. Among them are:

• setup of call center that's available from 8AM to 8PM
• increase in the number of Ceibal Móvils (the mobile repair-teams that offer repair services at schools around the country)
• reducing the price for repairs that aren't covered by the warranty (e.g. a keyboard replacement currently costs ~$10 and a display replacement ~$20)
• decentralization of repair facilities
• workshops focused on how to take care of XOs

It remains to be seen whether these measures have the desired effects but reading things like this blog entry (talking about XOs being blocked by the security system due to them not being able to access schools' WiFi networks and hence unable to renew their activations leases) and speaking to some people it became clear that more needs to be done. For one the call center hotline is only free to call from landlines but since a lot of Uruguayans use mobile phones (from which the call isn't free) with the combination of having to wait in line for several minutes is seen as making it unnecessarily hard to get the repair process started. When I inquired about this with Plan Ceibal I was told that starting in August calls from mobile phones would also be free and that efforts had been undertaken to have more calls answered in less than 30 seconds. Another thing they're working on is to display the call center's number during the boot-up of the XOs which I personally consider to be a great idea.

Additionally some of the rules that govern whether parents need to pay for the repair of an XO or not are rather arbitrary. One example is that keyboards which have more than 10 keys missing have to paid for to be replaced. Yet as I know from personal experience it only takes very little time (as in a couple of days) to go from a keyboard with one missing key to one where a dozen keys are missing.

The fact that only about one in three broken XOs is currently undergoing repair clearly indicates that there's a significant barrier to entry to that repair process. This is particularly surprising given that the repairs themselves seem to be handled very efficiently with Plan Ceibal claiming that all machines sent to their repair centers are returned to the owners in no later than 5 days.

More importantly however the majority of the measures described in the report relate to the repair process when really more attention should be paid to ensuring that the XOs don't get broken to begin with. This strikes me as somewhat odd as there is a clear understanding, both inside as well as outside Plan Ceibal, that the social environment is a key factor when it comes to the breakage of XOs.

Realizing this some of the RAP Ceibal groups have started organizing information events for parents before their children receive XOs. I had the chance to attend one of these meetings and the topic of how to take care of the machine as well as emphasizing that this was a tool for learning and not a toy was greatly emphasized. The volunteers I spoke to estimated that in the past similar talks seemed to half the number of broken XOs in some of the communities they worked with. Yet it seems clear that these volunteer efforts can only go so far. Plan Ceibal has apparently started work in this field as some particularly critical schools now receive visits from a team consisting of technical as well as educational staff. However I do feel that a more comprehensive outreach to all parents and teachers should be very high on the agenda of Plan Ceibal.

I will definitely keep an eye on this topic to see how things in Uruguay develop over the course of the coming months as well as speak to people in Paraguay and Peru about their experiences when it comes to XO breakage.

Update (14-10-2010): In the original version of this article I raised some questions about just how representative the data sample CITS' report - and subsequently the conclusions in this article - is based on really was. It was now pointed out to me that CITS' report actually contains more detailed information about the data sample (275 groups of pupils in 55 schools which are representative of 97,5% of the country's pupils) and that the study was conducted by the Statistics Institute of Uruguay's Universidad de la República. After reviewing this information I'm happy to drop my original concerns and now feel confident that this data is indeed representative of the nation-wide XO breakage rates as of April 2010.