CCK09: How to look at a blogroom (3)

(Versione italiana)

Abstract, A kind of introduction

Beginning to look at numbers

After the kind of introduction it would be natural to give some details about methods and tools but  I prefer to further clarify what we are about first.

I’m reporting data relative to an online class of 21 students who were teachers of primary and secondary schools from all over Italy. The class was about using digital technologies in teaching activities and took part of a three-years graduation in an all-online university called Italian University Line based on the cooperation of five italian universities

and a National Agency devoted to the research and development in education: Agenzia Nazionale per lo Sviluppo dell’autonomia Scolastica .

This is the class where the blogroom method was most successful. The students reported this experience as a very rewarding one and all had the feeling to have learned a lot of things in a meaningful way. Furthermore, they felt to be engaged in a community.

The impression of having learned a lot of things is remarkable because the course was organized in a very loose way. I did not declare a “program” as a list of contents but I suggested a number of objectives, partly to be determined in function of the experiences and the needs of students. I claimed that contents would emerge from activities and that the whole course had to be thought of as a path, not necessarily the same path for all the students.

The case of this class was advantaged by the small number of students with respect to the majority of my classes which may reach more than 200 students. However there were also strong adverse factors, such as the busy life of adult students and the 100% online nature of the course, that made the experience significant.

The thesis I’m anticipating here is that, the perception of meaningfulness claimed by the students has to be advocated to the  emergence of a living community, very much akin to a community of practice, a context where the learning process is substantially improved.

The idea is therefore to use this class as a kind of reference to be compared with other classes, in order to understand the most significant factors affecting the learning experience.

The formal part of this course consisted in eight assignments where the students were required to write a short essay or to comment about a specific activity. Therefore, in the conventional perspective, each student was supposed to produce eight posts on its blog, meaning a total of 168 posts for all the students. Instead, the students, wrote 484 posts meaning that 316 posts have been written spontaneously. In other words, we can say that the students wrote about 23 posts each instead of eight.

One could wonder how much pertinent these extra 15 posts per student are. Here comes another reason I have chosen this class among many others. These students are in the range 30-50, therefore they usually have families duties and a job. That is to say that these are highly motivated people and, actually, almost all the posts concerned the topics of the course. Having said that, I do not mean that students are blamed for broadening the subjects of their posts, and very often their blogs are quite reach and colored, indeed. However, for what I’m trying to show here, it is good to know that the extra posts had a high probability of being “meaningful”.

So far, it appears that those 21 people did a great deal of work producing a 200% more of what it was expected. Even if this statement has some value and can be used for comparison with other classes, it is worthwhile to look at the results in more detail.

First of all, the average result may be a significant estimator of the global performance of a class but it is a very poor estimator of individual performances that are by no means homogeneous. Let us look at the distribution of number of posts per student. In reality, the figure shows the number of posts for each blog. In all cases this corresponds to the number of posts per student except in one case were two students shared one blog.  The point is the first one at the left. Keeping this exception in mind, in what follows I will refer to the number of posts per student.

f1_plot_IUL_esplicitaWe see that there are very large differences among the students. As a matter of fact, we found that the 484 posts were distributed among 143 “due posts” and 341 “extra posts”. This is due to the fact that the five students at the right wrote less then eight posts for a number of reasons: some had still to finish at the time of data uptake and some others needed a reduced number of credits for this course. Fifteen students wrote more than the minimum of eight posts, and about half wrote more the double of the required posts.

The 70% figure of “extra posts” (341 over 484) can be thought of as a measure of the community effect, in some way. In fact, it is well known that when a population is just free to undertake something or to contribute actively, the so called 1-9-90 rule holds true. That is over 100 people, one starts contributing, 9 follow the first one and the remaining 90 just look at what’s going on. These last people are the so called lurkers in the Internet. According to such a rule, the plot should be quite different, showing perhaps one or two outstanding contributors and the rest with just eight posts. The plot should therefore be completely flat except for a very narrow peak at the extreme left, instead of being almost triangular.

However, even if the distribution of posts gives an idea of the contributions exceeding the required minimum, we still miss the most important side, which is represented by comments exchanges among the students.  In order to get a visual feeling of such kind of activity it is useful to look at the sociogram of the blogroom.f1_gplot_IUL_esplicita

The sociogram shows the connections among the participants as well as their role. The existence of a line between two nodes means that at least one of the participants relative to that nodes made a comment to the other but it may also mean that multiple comments in both directions were made. Red nodes represents students belonging to the class, blue nodes students from other classes, green nodes educators external to the class that got involved in the blogroom, the light blue one is me, the teacher. The layout of the sociogram was determined by means of an algorithm called “Multidimensional Scaling (MDS)”, which is capable to discriminate somewhat the role of different actors in the network solely on the base the connections.

Actually, the MDS algorithm did a pretty nice job by placing the most active actors to the right and the less active ones towards the left side. However, the most relevant point is the amount of connections among the nodes and, in this respect, it is worthwhile to compare this sociogram with the extreme simulations I  showed in the introduction, were a conventional class and an almost all-connected classes were represented.

The star-like sociogram describing a conventional class is compatible with the plot of the number of posts. At most, we cold add arrows towards the teacher to say that the students produced a significant amount of contents and that this fact made a difference for the teacher.

This represents a conventional were the students produce such a lot of information that the teacher begins to notice it …

This diagram represents a conventional class were the students produced a lot of content, beyond what was required by the teacher.

We could call this a kind of high-performance class with highly performant students but there isn’t any “network effect”, everybody is just working harder, both the students and the teacher. Instead, in our blogroom a great deal of communication took place among the members.

The sociogram of our blogroom gives a spatial perception of the web of connections but it is interesting to go further by quantifying those connections. We can do this by plotting the number of comments that have been made or have been received and, eventually, some measures of the role played by different members in the network.

f2_plot_IUL_esplicitaThe first twenty points at the left represents students belonging to the class and are the same of the precedent plot about the number of posts. The nine points at the right side are relative to students belonging to other classes or educators that engaged spontaneously in the blogroom. The last point at the right is the teacher. The number of posts written by these last members (black line) is not plotted.

The black line corresponds exactly to that of the precedent plot. The difference is that here are reported data for all the members of the blogroom and not only the students, such as students not belonging to this class or educators that became interested in the activities of this blogroom. For these new members the number of posts in their blogs is not reported here so that the black line drops to zero towards right.

The red line shows the number of comments received from other members by each student. In the jargon of social network analysis this is the “indegree” value of a node. The green line shows the comments made to other members blogs by each student, that is the “outdegree” value. Indegree and outdegree are measures of  “centrality” of a node, that is the relative importance of that node in the network. In our case, a high indegree describes a prestigious actor and a high outdegree an outgoing one.

There is a rather loose correlation among the number of posts, the number of received comments and that of written comments. There are people that wrote quite a large number of posts but experienced few contacts with respect to others who wrote a comparable quantity of posts. Conversely, there are someone who preferred to comment others blog instead of writing on its own. Such particular behaviours are worthwhile of careful consideration because they may reveal interesting approaches as well as potential problems. It is difficult to follow all the interactions that are going on within the blogroom, thus this kind of representation may be useful detect relevant situations.

Even more interesting is the quantity plotted as a blue line: the so called betwenness, another measure of centrality which takes into account how much a node “is between” other nodes. The following picture taken from the article Centrality of Wikipedia (in this article you can find also the rigorous definition of betwenness centrality) gives an intuitive idea:

Color (from red=0 to blue=max) shows the node betweenness. Image downloaded from article <a href=

Color (from red=0 to blue=max) shows the node betweenness.

Together with the degree measures of centrality, betwenness is among the “classic” parameters of social network analysis since it was introduced in the seventies, however it is still considered a very effective measure of centrality. The difference with respect to the degree measures of centrality is that these are local, counting how many nodes are connected to a node, whereas betwenness measures how many paths among couple of nodes a node is intercepting, were the nodes may be also many steps apart.

That is, betwenness measures how much a node is connecting different parts of the network and not only neighbour nodes. Actually, I found that this centrality measure was very good at detecting the most active students.

A node with high betwenness acts as a sort of catalyst enabling the spread of  a 1-9-90 kind of participation towards a broader distribution. Thus, the high-betwenness nodes may play a crucial role for the blogroom life. We can think to that nodes as “key nodes” to which it is worthwhile to pay attention in order to facilitate the rise of the community.

It is not to say that those students associated to “key nodes” necessarily will get high grades. This may happen but it is not the point. It is to say that the teacher should spend some time in facilitating the action of “key nodes” because this may foster the health of the whole network improving, in turn, the participation of all its members.

I foresee to use betwenness in much larger classes, where it is difficult to become aware of all potential “key nodes” in time. I would like to detect those nodes at an early stage so as to exploit their potential before the class is going to finish.

The instrumentarium of social network analysis is extremely reach and with the simple examples I have made we have only scratched the set of possibilities. Of course, abundance of possibilities does not mean certainty of success but it is worthwhile to explore them provided we stick to this objective: devising methods to nourish the community in the blogroom giving rise to an experience of meaningfulness in its members.

In successive posts I will tell about other ways to look at blogrooms and as well as about the methods I’m applying.


Inserisci i tuoi dati qui sotto o clicca su un'icona per effettuare l'accesso:


Stai commentando usando il tuo account Chiudi sessione / Modifica )

Foto Twitter

Stai commentando usando il tuo account Twitter. Chiudi sessione / Modifica )

Foto di Facebook

Stai commentando usando il tuo account Facebook. Chiudi sessione / Modifica )

Google+ photo

Stai commentando usando il tuo account Google+. Chiudi sessione / Modifica )

Connessione a %s...