Data from the course #Lak11 can easily be misinterpreted when personal interpretations are made.

I find myself far down the data - http://ariadne.cs.kuleuven.be/monitorwidget-lak11/. It is true that I have not spent so much time in Moodle, not written so much and doesn´t read so many posts. My ambition when I started the course was to be an active participant, but due to time constraints and above all the fact that it was a completely new area - I have not been as "visibly active" as I hoped to be. I became a lurker in the course.

I'm thinking about what it means that I am found far down in this data. It shows quite clearly that I have not been as active in Moodle, however I have learned a lot. I have been following the lectures and seminars, read the literature and related literature, read blogs an so on. None of this is shown here. The data do not show what I feel - that I have learned a lot. It is true that the data does not show anything more than that I have not been so active in this forum, but it's very easy to feel that my efforts were not so great. It is also easy to interpret the information in the way that I have not learned so much. In any case, that is what I feel when I see the data and make comparisons with the others.

Fortunately, I am aware of what the data says and know about what you can use the information to. However, it is probably not as much knowledge about this among teachers and students. Information like this is still often not available to students and teachers, but it's probably safe to say that similar information will become common in the coming years. Then will it be vital that there is a high level of knowledge among teachers and students on how information can be interpreted. It may not be used because it is easy to measure and compare. Students and teachers emotional interpretations, as I did, must be prevented and the only way is that teachers and students learn more about learning analytics.

Thank you for a rewarding course.
I look forward to the conference as a lurker, but I´m sure that I will learn a lot and that it will appear in any data.

Thanks Tony!

After following Tony's great instructions, I managed to get "my facebook network. "
It is with some hesitation I show the visualization of my "Facebook network", because it says a lot about me even thought I'm not a frequent user of facebook. The goal to visualize relationships in Diigo groups have come a step closer, next step is to continue to read Tony's tutorial.

Reflections on Bakers presentation week 2

In previous post I wrote that I think that the structure of this MOOC is brilliant, but the content is also impressive!

There have been two presentations so far in the course, both were very valuable, for me. What I particular fell for in the last presentation by Ryan S.J. Baker about Educational data mining (EDM) was:

• Unfortunately is the most interesting parts in the presentation not useful for me, when Baker talked about “Gaming the system”, this is extremely relevant and I think it´ is the very hard for the teacher to observe.
  
• PSLC Data shop is interesting, but I´m asking myself what does the students learn from the program that was used in the examples in the presentation. 
  
• When Baker tried to explain differences between EDM and Learning analytics (LA) and started with that there are many similarities between the two communities.
   
  
• The ethical questions around data mining are extremely important. Do the students know that the data is collected?

There were some very interesting posts in the chat:

• Do the students know that this is happening?
  
• Who is this data for?
  
• EDM could answer Who, When, Where, but cannot answer Why.  Are there studies connected to complement the findings of the data analysis with a more Qualitative research that could explain the behaviour or propose a solution?

• “Does EDM use naturally occurring data and LAK includes how to generate the data?”

• “For me the difference between EDM and LA is that Analytics should concentrate more on how the information extracted from the students actions is used to improve learning”

• “LAK doesn't necessarily imply data mining - can build an explicit user model for instance from known inputs designed for that purpose”

The definitions of EMD and LA done during the presentation were clarifying:

Bakers definition of EDM

“The area of scientific inquiry centered around the development of methods for exploring the uniique kinds of data that come from the educational settings, and using those methods to better understand students and the settings which they learn in.” Baker 2010

Siemens definition of LA "Learning analytics is the measurement, collection, analysis and reporting of data about learners and their contexts, for purposes of understanding and optimising learning and the environments in which it occurs"

From my point of view "big data" (discussed in the presentation) is interesting and can be used to design better and other tools and also improve education, but this will be in another scale than the teacher. Probably will not the teachers use big data, rather system administrators, product developers,  researchers and so on. 

What I see as the largest question, except ethical questions, is how the analysis of the gathered data are made available to students and whether it can be created built-in feedback mechanisms in software. 

Today I can see that EDM can be used alone to answer questions about students learning, that it must be combined with qualitative research.

Copy the structure of a MOOC to design of education in upper secondary schools?

It´s overwhelming to join the MOOC course #LAK11, first of all the structure itself. In this course Siemens use Netvibes to aggreagete the information from the participants, google groups to send out Daily mails, moodle as so on. I think it´s an brilliant solution.

Is a similar solution possible to use instead of a LMS in a class in upper secondary schools? What would happen if that is the structure in all classes? What would the situation for the students look like? Is it worth to put a lot of time to learn the students to aggregate the information? Are schools and teachers willing to let the students take that responsibility?

What kind of information, and communication must be hold in closed groups?

These are some, for me, unanswered questions, that I would like to have answer to.

But there more questions. What are the “learning community” is it the class and the teacher. Of course is the learning community bigger, if the we look from the student perspective. Students communicate with friends, parents and so on, but this is usaly not visible for the teacher or other students. I dosen´t make sense to structure the communication between students, teacher in a similar way as #LAK11, if the idea is that the communication should be closed in the class.

I have seen several examples where there is a problem that the students and teacher not are able to communicate with other outside the class in the learning situations through the LMS, but does the teacher and students want to open up the “classroom” to let other take part in the communication?

           

Where, when, why students learn outside the teachers scaffolding is not anything the teachers (schools) can decide. But by using the same tools (and also to show the possibilities to learn with them) for communication in the education as the students use outside the “formal learning” can make off course open up the “learning community” and make the connections more visable. It will also give possibilities for teacher and other students to take part of what the students learn outside the classroom.

The “extended learning community” will never be able to see, but it can be more

How should an upper secondary school be organized if the education is based on the ideas - personal learning environment (and network) and the structure of a MOOC?

The structure of a MOOC can´t be the fundament of the education structure, but some parts would be very useful to use.

Reflections on week 1

The literature is largely on the analysis of large data sets and how institutions can see patterns in students' use of LMS. The term academic analysis refers to how institutions gather, analyze, and use data (Elias 2011). Bakers article The State of Educational Data Miningmakes is a  valuable “review of the history and current trends in the field of Educational Data Mining (EDM)” (Baker and Yacef 2009) , but it not easy to enter a new area and understand the article. On the other hand am I sure that the content in this article will become clear to me during the course.

Before this course I saw the potentials of data visualization in the way it´s been used in Knowledge Forum where the technique Latent Semantic Analysis were used (Teplovs and Scardamalia 2007). They use the visualization tools with the goal “to extend this analytic framework to enable embedded and transformative assessment of the knowledge building process, not simply assessing its end point.” Their goal requires that they present the results (from the database that store information such as time-stamped information about activities and interactions and knowledge objects and artifacts) to inform the users about their ongoing process that answers questions such as “in what contexts has a particular individual worked? What are the dominant ideas in the discourse space?” (Teplovs and Scardamalia 2007)

Their work is interesting and particulary the use of the information to formative assess the users, where formative assessment “can be conceptualized as a cybernetic system with feedback loops serving to drive the system in new directions”  (Teplovs et al. 2007)

There are several different tools; Participation and Collaboration Tools, Writing Analysis Tools, Semantic Analysis Tools and “The Participation, Collaboration and Writing Analysis tools focus on surface features of contributions. The Semantic Analysis Tools deal with the meaning of the discourse.” (2007)

The function of the Social Network Analysis Tool, “which displays the social relationships among participants based on patterns of behavior” (Teplovs et al. 2007) seams to me be easiest to understand and use, but in other tools, but also “The Semantic Analysis Tools deal with the meaning of the discourse.” make sense to me.

I´m not for the moment using Knowledge forum but the functions of the tools that are used are interesting. I would like to be able to analyze the social relationships when the students use different web-tools to see who the students are communicating with. So the first step to enter this area is to visualize students networks and it is striking that this seams to me be the historical development. Baker writes that relationship mining that was dominant between 1995 and 2005, in 2008-2009 it slipped to fifth place, (Baker and Yacef 2009)

I´m not sure that social network analysis is the same as relationship mining and I can´t place it in Bakers classification (Baker and Yacef 2009) of educational data mining methods.

• Prediction
•          o Classification
•          o Regression
•          o Density estimation        
• Clustering
• Relationship mining                         
•          o       Association rule mining
•          o       Correlation mining
•          o       Sequential pattern mining
•          o       Causal data mining
• Distillation of data for human judgment
• Discovery with models

I can see the opportunities with learning analytics, but there are several concepts that I have to understand to get more engaged in the course.

I also have to learn how to mine data from different web-tools and create and manage databases outside LMS/CMS.

There are also some more readings that seams to be very interesting:

Dawson, S. (2008). A study of the relationship between student social networks and sense of community, Educational Technology and Society 11(3), pp. 224-38.

Dawson, S. (2010). “Seeing the learning community: An exploration of the development of a resource for monitoring online student networking. British Journal of Educational Technology, 41(5), 736-752. doi:10.1111/j.1467-8535.2009.00970.x.

Dawson, S., Heathcote, L. and Poole, G. (2010). Harnessing ICT potential: The adoption and analysis of ICT systems for enhancing the student learning experience, International Journal of Educational Management 24(2) pp. 116-128.   

Finally I will end this with the most interesting questions that I have found in the blog posts so far.

Dr. Kelly Edmonds raised some important questions

“Can we gather data about student attitudes and interests through social networking sites, their interaction in online learning platforms, their use of online library systems, etc. and blend the data to discover who they are, what they want, and how they learn along with their progress?

This leads to my second reaction: how will we be able to logistically mix the variables and constructs of the various data types? Will we be able to find a way to compare student preferences (from likes and dislikes on Facebook), with chosen modes and frequency of communication (email, LMS, smartphones), with grades and course evaluations?”

Reference:

Baker, R. S.J.D, and K. Yacef. 2009. “The state of educational data mining in 2009: A review and future visions.” Journal of Educational Data Mining 1:3–17.

Elias, Tanya. 2011. “Learning Analytics: Definitions, Processes and Potential.” learninganalytics.net/LearningAnalyticsDefinitionsProcessesPotential.pdf (Accessed January 18, 2011).

Teplovs, C., Z. Donoahue, M. Scardamalia, and D. Philip. 2007. “Tools for concurrent, embedded, and transformative assessment of knowledge building processes and progress.” Pp. 721–723 in Proceedings of the 8th iternational conference on Computer supported collaborative learning.

Teplovs, C., and M. Scardamalia. 2007. “Visualizations for knowledge building assessment.” in AgileViz workshop, CSCL.

 

Lak11 week 1

Last week was an intensive week, which means that it was no time for participating in the MOOC #Lak11. To start a MOOC like this is more difficult than missing the start in a traditional course – the feeling is that it is impossible to read enough to catch up.  But I´m shoure that this isn´t just because I missed the start – it is about the structure in a MOOC.  I start to accept that I have to pic what I find interesting and not try to read everything – (and listened to Simens recommendations).

As totally new to the field educational data mining, it will be some work that have to be done to understand the mathematical methods and also the tools that are possible to use. But the big challenge is to learn more about the security aspects. What information is a teacher allowed to collect and what information can be collected, anlysized and published from an ethic perspective?

According to my experiences as a teacher. Teachers have knowledge about different groupings in the class and also work with the class to create new "learning-groups", to dived the class in different ways that to create a whole “learning community” of the class. It is very important  that students meet each other and get used to learn (in dialogue process the information) with different classmates.

When working with web-tools is it much more difficult to get the knowledge (“feeling”) about who the students responds to (in dialogue process the information). Of course is it possible to dived the class in groups that work together in web-tools. But when the students for example start the work with personal inquiries in web-tools such as Ning and the students who are interested in the same topic responds on each other comments, it is much more difficult to get the knowledge (“feeling”) about different sub-groups in the class. Here is semantic (social) network analyze probably very useful to see different and get knowledge about the groupings in the class. This is seams to important for teachers to understand to be able to work with the social structures in the communication in web-tools

Looking forward to Learning Analytics #LAK11

I´m looking forward to the course #LAK11 and hope too learn more about semantic (social) network analysis in educational design. I´ve done some first steps in analyzing students activity in Diigo-groups. 

Here are two visualizations from relationships between tags in a Diigo group. The folksonomy (tag cloud) was created by students in upper secondary school in a Biology class. 

The data were processed with a forced based algorithm and in figure 2 also Louvain method are used to find communities (groups).

---

Figure 1

Figure 2

I especially hope to learn more about data mining and tools for data mining in the course. To facilitate the work with collecting data. If you already have some tips for me, please write a comment.