Do Children Want to Learn with Sugar?

olpc walter bender
Walter looking for Sugar Labs answers

IT Wire reporter, Sam Varghese, recently wrote about how Sugar Labs tries to pick up the pieces now that its fully spun out of One Laptop Per Child. In his article, he chose to highlight a question posed to Walter Bender at LCA2009, Australia's premier Linux conference:

When Martin Krafft, a senior Debian developer, asked him, "have you asked the kids whether they want to learn this way?" Bender danced around the question. When I reminded him that he had not answered Krafft, he ducked once again. A third attempt to ask him was nipped in the bud by the conference organisers who announced that the time limit for his talk had been reached.

What an interesting yet foolish question. It supposes that children have the faculties to assess the learning performance of competing pedagogy and the software that supports it. Kinda hard when you're a 5 year old. I think better questions, all still unanswered are:

  1. Does computer technology increase learning outcomes?
  2. Is that increase greater than other options for the same resource investment?
  3. Of computer technology options, does Sugar help children learn better than others?

Each of these has a greater impact on the trajectory of the Sugar Learning Platform, and the use of computer technology in education in general, than posing impossible questions to Walter.

Oh and while we're asking Walter questions, what questions about Sugar do you have?

Related Entries


Very nice article Wayan. I have to say that OLPC News is a great source of information and has developed into a must go site over time (I have followed it from its early days).

Regarding Sugar I would just ask what does it do for the teacher?

I was teaching English in rural China a few years back, and I doubt not that there's a lot of value in the OLPC project, and that Sugar is a good idea.

But I think that it lacks pragmatisms in too many way - and contains too much dogma.

In my limited experience, the smarter kids in the class will figure out whatever User Interface you throw at them. The real value of Sugar would be if virtually all of the kids in a given class felt more comfortable using Sugar than a UI like GNOME or XP. That would make for an interesting study.

Wayan - I agree with the previous post...very nice article.
When I started a trial eLearning program for disadvantaged students all over California I had to do extensive assessments to show progress (under the guidelines of NCLB)
I did not use Sugar as it was not available then but I did learn that most disadvantaged students were not able to use traditional computers systems and I did need to custom make the interface easier to navigate. I only ran the programs the last two years but here is the data we collected.

In our traditional in-person program we saw average academic improvements of 15%, but in the online program where the students took classes virtually through computers we provided them we saw an average increase of 38% !!!

There are a number of factors but can contribute to the differences and accuracy but it was very obvious that the students who wanted to participate were more engaged in the learning because of the inclusion of the technology.

On the flip side we did see a lower attendance rate because there was nobody to make the students log-in and participate.. if their was a tv show or soccer game they would rather be involved in... learning would take the back seat.

Not every pedogogy will work for every student, and while new and exciting technology based learning tools are coming out but they need to be tempered with traditional learning methods.

"have you asked the kids whether they want to learn this way?"

But we know from the deployments that the children love Sugar.

That said, it is right that it still needs to be determined empirically whether or not it actually helps them learn.

The question has great validity; the fact that the Sugar team cannot speak to the role that children played in the design of the interface (either as subjects in usability studies or as partners in design) says that, practically speaking, Sugar is no more "kid friendly" than GNOME or any other interface.

Consider the work of Druin et al. in the Human Computer Interaction Lab at the U. of Maryland. Their paper "Designing a Digital Library for Young Children: An Intergenerational Partnership" from JCDL 2001 is an excellent example of what a child-centric design process would look like.

If you really want to design an interface for children, you should involve children in the design process. It isn't really debatable.

The questions put to Walter were not impossible---they were spot on, as the question(s) put the user of the interface first. If Sugar was developed in close conjunction with a group of 7-11 year-old school children, then Walter could have said "Yes, it is good for children (and they like it) because our user studies demonstrate this is good."

Matt wrote:

"The question has great validity; the fact that the Sugar team cannot speak to the role that children played in the design of the interface (either as subjects in usability studies or as partners in design) says that, practically speaking, Sugar is no more 'kid friendly' than GNOME or any other interface."

Absolutely correct, Matt.

Sam varghese speaks from his ass. Everyone knows this. No wonder his write ups stink.

Even though the article raises an important issue, I can stop wondering... Are the kids ever asked if they want to learn the way they do? specially those in developing countries?

its not a foolish question at all...While i agree that directly/literally asking kids if they want to learn "this way" (aka using sugar) is silly, a realistic idea (which has been talked about on this site a bit) is to go to a school with a large number of children around the age of 5, and use a seires of methods to determine what would be the best lay out. could be done by asking them questions, could be done by assessing how easy it is for them to use them (aka tell them how to do something once or twice and then see if they can do it again by themselves); and/or, instead of asking them what changes they would like to see, just try different setups to see which one the kids overall take to the best.