Thursday, April 19, 2007

D.Watts on Social influence and popular songs

This study is pretty interesting: there were 14,000 participants who were asked to listen and rate songs by bands they had never heard of. The point was to study the social influence, i.e. how seeing cues from other people, like Top10 downloads, no of downloads, etc. would influence on people's choice.

The set-up of this study is pretty neat, the participants were sliced into eight parallel “worlds” so that participants could see the prior downloads of people only in their own world. Everyone started from the same line, zero downloads — but because the "worlds" were kept separate, they subsequently evolved independently of one another.
What we found....In all the social-influence worlds, the most popular songs were much more popular (and the least popular songs were less popular) than in the independent condition. At the same time, however, the particular songs that became hits were different in different worlds, just as cumulative-advantage theory would predict. Introducing social influence into human decision making, in other words, didn’t just make the hits bigger; it also made them more unpredictable.

Our experimental design has three advantages over both theoretical models and observational studies. (i) The popularity of a song in the independent condition (measured by market share or market rank) provides a natural measure of the song's quality, capturing both its innate characteristics and the existing preferences of the participant population. (ii) By comparing outcomes in the independent and social influence conditions, we can directly observe the effects of social influence both at the individual and collective level. (iii) We can explicitly create multiple, parallel histories, each of which can evolve independently. By studying a range of possible outcomes rather than just one, we can measure inherent unpredictability: the extent to which two worlds with identical songs, identical initial conditions, and indistinguishable populations generate different outcomes. In the presence of inherent unpredictability, no measure of quality can precisely predict success in any particular realization of the process.

This makes me want to test and set up experiments, too. In the project that I'm part of, and where I will get my data, we are planning some experiences on the input part of the tags to see how social influence in terms of seeing other users' tags when inserting own ones, will effect on the nature of tags, their number, their convergence, etc.

But, on the retrieval side of things this would be very interesting too! To have two different interfaces to see the search result list, where on the one there would be all the social cues for social influence (no of downloads, no of bookmarks, others' tags), and on the other one there would be nothing. The experiment would test whether the users, in this case teachers, would be viewing the metadata of similar resources and what would they actually download, bookmark and rate, if they did any.

Well, actually the latter is the situation as it is now. So maybe I can just compare the data from this year and the year after, when we actually start implementing the social navigation part.

In NYTimes

Science 10 February 2006:
Vol. 311. no. 5762, pp. 854 - 856
DOI: 10.1126/science.1121066

Supporting material:

Thursday, April 12, 2007

What tasks teachers have on a learning resources portal?

The attitude of "if we build it, they will come" has resulted in national and regional learning resources portals where the offer, no matter how many learning objects or assests, does not necessary match the need of teachers. Why is that? What is it that teachers look for?

Curriculum coverage

I first started by looking at 29 different learning resources portals that national and regional educational authorities offer for K-12 teachers in Europe. My task was to find out how many of them offer curriculum related material, i.e. so that a teacher, knowing that tomorrow he has to teach an area that covers a certain goals of the national curriculum, can just go to the portal and pick a resource that actually goes through this particular area.

This type of standards-based curriculum material seem to have been on the offer in the US for some time. Two examples could be the DLESE, focusing on Earth Science and IDEAS, a repository held by Wisconsin educators. In both teachers can search for a curriculum coverage.

I found that in about 10 out of 29 examples in Europe teachers can explicitly look for curriculum coverage on learning resources, however, it was not always very clearly indicated which goals or skills a resource aims to attain.

On rest of the portals teachers were able to find resources that were categorised by the school level and subject, thus, by using teachers internal knowledge of curriculum, they would, with little poking around, find the matching material. But the problem with this case is that there is nothing that makes the teachers' tacit knowledge externalised for other users, no one else can take advantage of the fact that this teacher knows how well this piece of learning material covers certain areas and goals of the local curriculum.

In MELT, we are trying to get to the core of this problem by encouraging teachers to add tags to resources that they find in the repository. We'll be eager to see whether tags can externalise any of that tacit knowledge that teachers have, and that could help other teachers in finding and using the material.

In Calibrate, another project that I have a minor involvement, we have an opposite approach to the problem. A few topic areas have been selected where 3 EU-countries share a similar curriculum. We try to map the different curriculum through its goals and skills, and match that with the material available. A complex task!

Tasks at hand

Then, I started thinking whether learning material that clearly covers the standard-curriculum is what teachers want? One could think that a busy teacher would really appreciate it, however, there might be other reasons why do they come to a learning portal. So I made a little poll with some tasks that I thought teachers could have, and asked some teachers to vote. I only got 36 votes so far (you can still vote here), but it seems that, at least for these teachers, the curriculum coverage was not what they preferred!

The majority of the respondent teachers (I don't know who they are) seem to be leisurely window-shopping while at the learning resources portal (I browse around to see what is available) or looking for material that could support the lesson that they are planning. Only one teacher is looking for learning resources with an exact curriculum coverage!

This rises two questions in my mind: either teachers have given up to look for curriculum covered material, because a) it never was on the offer, b) it was never successfully on the offer or c) they have better sources for that, like school books, etc.

Secondly, maybe teachers don't really know what to expect from a learning resources portal or a repository. Maybe it was never clear for them what was the intended goal of a learning resources portal and they just come there to see what's up, what's in there and maybe they return if something good is found.

Seriously thinking, do we even know for what tasks and goals the resources repositories are build? If we think of libraries, we know they have a clear goal, or a school book, yes, it has a clear goal. But a LOR (for learning object repository), isn't it something in-between, kind of pretends to be one or the other, without being either of them.

I just looked at the sneak-preview for Yahoo!'s new service for teachers. Pretty neat. They clearly aim it to be to create learning resources, and re-use the ones from other teachers. They also offer neat peer-networking possibilities. It's about using and re-using material that is in the center, not searching for it! A very different focus.

So, I think what should be build in on learning portals is a better support for the tasks that teachers and learners have at hand. So, when they come to a portal,
  • if they clearly are just window-shopping, let's provide them with first-grade Champs-Elysees shop-windows. Make nice pre-views available of resources that are there with added value lesson plans and case studies how others have used them in their lessons. Allow browsing other users' collections of learning resources with annotations and comments. These other users can be from any part of the continent, as the goal is to inspire and show how things can be done. If we know anything about the user, let's try to match them with like-minded peers!

  • if they look for material with curriculum coverage, let's lead them to an area where they can either search for material with curriculum coverage, or browse bookmarks and tagged resources for cues from other teachers on attaining certain skills and goals. For the latter, it could be useful to first show "traces", e.g. bookmarks, tags, pedaggical annotations, from teachers who come from the same area, like Yahoo! peer-network allows getting close to teachers from the same area (=same curriculum).

  • etc.
The point that I'm making is to be clear about the task, and the information seeking patterns in general that teachers have, and then match it with the best way providing search, social navigation, recommendations, shop windows, etc., but always thinking what would yield the benefits for the user and task at hand.

That's something to study deeper, and don't worry, I'm on it ;) We don't know yet if the best benefits can stem from using underlying social networks (location) or more implicit ones (profile, tags, similar bookmarks), or from using a search or what?

Wednesday, April 11, 2007

Social Navigation as seen a decade ago

I'm always fascinated when I find "old" papers that still resonate today. Well, I'm not talking about the manuscripts from the Library of Alexandria, but I just came across a paper on "Design Principles for Social Navigation Tools", written in 1998. The principals still seem rather relevant!

That makes me think; what was social navigation about a decade ago and how differently we perceive it today, after all, there was no social bookmarking nor tags out there, like we know them today.

Defining different flavours of social navigation

Social navigation can happen in many different forms... One may distinguish between direct and indirect social navigation [Dieberger, 1998, Svensson, 1998]. In direct social navigation, we talk directly to other users. In indirect social navigation, we can see the traces of where people have gone through the space, as for example in the Footprints system [Wexelblat and Maes, 1998].

In my context of use, direct social navigation could be seen to happen through networks of friends and colleagues, that the users of a social tagging system have established. Or, it can also be sort of "ask the expert", or ask your colleague type of thing. Direct social navigation could most likely take place in sharing pedagogical practices, for example.

Indirect social navigation, on the other hand, would be like following other users collections of bookmarks, browsing them through " xx other user have this in their collection" or through common tags, for example in the personal or common tag cloud.

Furthermore, social navigation may be intended or unintended by the advice-giver.. distinction can be made for when the advice-giver is one particular person, known to us, or when it is just a group of anonymous users that have happened to navigate through the same space as us. In-between these two extremes, we may have groups of users that are similar to the navigator in terms of interests, profession, knowledge or task. The advice-giver may also be an agent [Foner, 1993]1.

The idea of having intended or unintended advice-giver is intriguing, many times in collaborative environments for learning, we are very occupied in setting up advisors whose intend it to give advices. But in the real world, people are somewhat shy asking a specific "advisor" for hints or help, peers seem to work better, or many people look for "traces" for that purpose. So, it seem to me that it is very important to design a lot of unintended advice-givers to help social navigation in learning and teaching contexts, they can be anonymous crowds, agents, traces, annotations, hints of task orientation, shared interests, or what ever we can think of.

I wonder if this table makes sense like it is now: (I'm not talking about blanc spaces...)

Unintended advice-giverIntended advice-giver
Indirect social navigationTag cloud guiding the navigationA recommendation (person, agent,recommender,..)
Direct social navigationBookmarks from network of friends to navigateA friend, expert, peer gives advice or recommendation

The 6 design principles

The principles according to Forsberg et al. (1998) are Integration, Trust, Presence, Privacy, Appropriateness and Personalisation. The examples are pretty hilarious, really like from 10 years ago, but nevertheless, the principles stand like they should!

It is emphasised that social navigation should be part of everyday tools to make best use of it. This is what we see nowadays a lot, more and more things are integrated in our workflow, for example the browser with all the add-ons and blug-ins has become a central tool.

To make the best use of social bookmarking, it is of utmost importance to make the process easy and part of what one was doing right at that moment. The delicious-bookmarklet added in the browser toolbar is a good example, it is so easy to use that you hardly even have to stop what you are doing at that moment to bookmark. Thus, you leave more traces for others, besides arranging your own information space.

The idea of Attention Metadata is another one, if you use Slogger to generate Attention Metadata on what all you do on your web-browser, you don't even have to think of doing it.

This refers to how do you make the presence of others shown to users, how do they know that someone has been here before. Annotations (tags, comments, evaluations, opinions, ratings...) are a perfect way to do that (Kilroy was here!), you know that someone passed through that space.

Or showing how many other users are online at the moment, think of how much fun is it to log into Skype at 1am and see that despite the quietness of your work room, some other people are out there still slaving away.

In order to take a note of someone else's doings, it helps if you know whether you can trust them, are they a reliable source, do they like the same things as you do, etc. When reading film critics, one quickly picks up the critic who is like-minded, and disregards the other one who always seem to have too much of a mainstream thinking. Similarly, to trust the source for online social navigation can become crucial, if there are many ways to go forward.

In some situations one design choice is more appropriate than the other one, for example indirect social navigation can be suitable for finding inspiring information, whereas you might rather turn to someone to talk to when you have a specific question at hand. This is in my opinion related to finding a means that fits the purpose of the information seeking task at hand, something that I've been mulling around a lot and is related to the Human-Computer Interaction framework.

This concerns the issues of making users aware of the traces that they leave, data that is logged and used for personalising their searches, etc. This is also related to some codes-of-conducts that some systems have. Very important for things like Attention Metadata, Google personalized, etc.

Personalising Navigation:
"Social navigation provides excellent opportunities for tailoring navigational advice to individual user's task, knowledge or abilities". How I see this, it is related again to the information seeking tasks, interests, etc.

Forsberg, Mattias and Höök, Kristina and Svensson, Martin (1998) Design Principals of Social Navigation. In: 4th ERCIM Workshop on User Interfaces for All, Stockholm, Sweden.

author = {Mattias Forsberg and Kristina Höök and Martin Svensson},
year = {1998},
title = {Design Principals of Social Navigation},
address = {Stockholm, Sweden},
url = {},
booktitle = {Proceedings of 4th ERCIM Workshop on User Interfaces for All}