In the last weeks I have been working in my spare time on a program that shows the complete corpus in one view and allows some filtering. After some experimentation I came up with a grid structure (12 * 13 = 156) with a additional navigation menu on the top. The navigation menu includes elements that lets you filter the graphics (linking and brushing).

I have not included all investigated dimensions and currently thinking about what to add, and how to do it. Really would like to incorporate some other visual representations the current ones are not mind blowing, but they are functional. I am thinking about a pictorial treemap. Thus, for instance, all graphics are coded on main news category. What if you make treemap out of that information, incorporate screenshots from the graphics in the map. The screenshots themselves are clickable and more information is provided in a information screen. It could be interesting to find out if it would work.

The usability aspect is momentarily not great. For instance, you can use the arrow keys, but without knowing that, no one would guess. The same thing counts for the filtering of the bar chart, how do you know you are allowed to click on a bar and how do you know that the image in the detail infobox is a link? Thus, need to think about some solutions, some are easier than others. The current version can be found online here (be aware, it’s a pure Flash project). And, like I said, just a prototype.

When you read articles regarding my topic you encounter a lot of different terms for, what seems to be, the same thing.

For example, there does not seem to be one word for this type of journalism. Some give it the general term ‘online journalism’ (Mark Deuze for example), Tim Berners Lee talks about ‘database journalism’, Simon Rogers about ‘data journalism’ and others call it things such as ‘computational journalism’, ‘multiple journalism’, ‘hybrid journalism’, ‘multimedia journalism’ etc etc.

I told a friend about my confusion with the names, she (a museum studies student) told me that by art studies an art period will only be named (e.g. ‘modernism’) after that period is over. During the moment no one knows for sure in what kind of we period we live in. Only by looking back we can  come to a consensus of a name. Maybe it is same with this, it is too new and broad to come to one name.

I think I will stick with online journalism for now. I wonder which term will eventually be the most common used one, my guess at the moment is on ‘data journalism’, but can’t be sure.

Some things have become clear in the last weeks. Something I knew a long time for a long time but never acknowledged: this project is too big. When I started with my thesis I knew it would be about visualization but did not have a specific topic. I first thought about writing about projects by Jonathan Harris, Aaron Koblin and Ben Fry.

After the infographics day in Zeist last year I decided to go for newspaper visualizations. At that time, Wikileaks was a big thing and, for a new media student, incredibly interesting. I also was intrigued by the talk Gert Nielsen gave about journalism, abstractification and aesthetics.

There are different reasons I choose for The Guardian and The New York Times, some are good other a little random. I think the most important reason was that I was familiar with the visualization work of both papers. And I had used the NYTimes API for a research regarding an EFF research concerning their main issues over time (which, surprisingly, is quite stable).

So, I decided to talk about visualization within those newspapers and about interactive narrativity. There is a real struggle between telling a story and having interaction, a discussion that reminisces the ludology/narrotology game studies debate from a couple years back.

But it was too big, and after getting quite lost in narratology/usability/game studies/human computer interaction (HCI) readings, I made a focus switch in November last year. The main focus would be about the two papers and about everything that encompasses this new visualization field.

If you want to do this, you have to talk about topics such as data, accountability, ethics (dirty datasets), web ephemerality, collaboration, form and function, the merging of separate fields, the journalist/programmer, template or custom design, web technologies, education, usability, interactivity, public technical savviness, visualization techniques and much, much more.

It is way too big.

I had started a couple months before to collect graphics, just to get an idea about what exactly is out there. It was really interesting to see how things changed in just a couple years, and how the graphics ‘evolve’.

But, as people have told me for a long time: this was not the way to go. And me, being stubborn as I often may be, did not listen. The thing is: I study at the humanities faculty and we don’t do statistics. We mostly do qualitative, literature-based research. A big, quantitative research is walking on the edge with your arms spread out and your eyes closed.

Which is frustrating.

To give a better context: I received my bachelors in information & interaction design at CMD, a study which could be best described as a combination of design, technology and usability. My teachers there love thinking out of the box, the highest praise you can get it when a teachers says that is very creative. It is about making things, experiment. You will often fail but sometimes make something quite good. When I graduated I received the book ‘Whatever you Think, Think the Opposite’ by Paul Arden. Which tells the whole story.

The study I follow University at Amsterdam is different: it’s academic. It is mostly about reading and writing: building on the works of others (which of course is also the case when you design something, but you can’t just make something up). It has to be well argumented, described and you always will have a long limitations list. It took me a while to settle in.

However, when I started with the thesis, you have a lot of freedom. I forgot some of the  things I had learned the last years and grappled it from a “out of the box thinking” perspective. It was a body-first thing. Which could be described as interesting, educational but mostly inefficient.

I haven’t been playing by the rules. Part of which I am grateful for, but momentarily it is not helping me to actually graduate: a thing I really would like to do.

Thus, to make a long story short: I have to make thing smaller. There has been last week a change within supervision and while my so-called leash was long, it’s now pulled in. This means that I have to throw away a lot of stuff, only focus on one main question and a couple sub questions.

Yesterday we made schedule, which encompasses the following two months. If things go all right, I will graduate in May. I hope a little sooner, but am not counting on it.

This is, again, frustrating. Especially because I receive a lot of questions of when I will be finished and too send the thesis when it is done. It feels like failure. But that’s the way it is. I work hard, finalized and postponed other projects, but can’t go faster than this. It is probably wiser to slow down.

Please keep in mind: this is only a master thesis from a student who started to read academic literature less than three years ago. I am not sure if it is going to be good, I hope so, but can only be certain when it is done. The good news is that both my previous as my current supervisor are still enthusiastic about it, I am extremely grateful for their support.

And you know what? Things are finally beginning to make sense now, I use one main theory from the open source community and it seems to be flexible enough to incorporate most things I would like to tell.And I have now have time to go to Malofiej without working 24/7.

When everything is done, I will put the excelsheet online (with the dimensions I am sure about it). It would be nice if others could use it for there own research and visualizations.

Either way, I think it is save to say, “curb your enthusiasm”, that’s what I am (trying) to do.

And if you tell something about visualization you can’t only use words, can you?

I decided to do something with the authors from the graphics in the corpus. Most graphics have a credit line, and one column in the excel sheet contains these names. So, with this idea I started to experiment with Processing code from the book Visualizing Data by Ben Fry.

I used this data to create two graph networks. The connections (edges) between names signifies collaboration. The size of the bubbles (nodes) represents the total of projects an author has worked on. The screenshots below show some of the results. It’s quite interesting to see the clustering. In both cases only four people are separated (by pairs). It displays the collaborative side of visualization.

The Guardian

The New York Times

I am looking for solutions to make the edges more meaningful. It would be interesting to see how often certain people collaborate with each other. My current solutions (line thickness, color) were either messy or unclear so have to work on that. And probably need to have a header, legend, stop the spastic movement of some of the nodes and a better node placement (the interactive version lets you drag and drop the nodes plus activate them).

It’s a big relief for me that the collection is forked. The page has received quite some visitors in the last weeks and it wasn’t really made for that. It currently loads 2.5mb the moment you open it for the first time. I plan to do site optimization (and some other things) in the future but it is not the main priority now.

This may be this is not the smartest thing to do from an academic perspective (people are generally quite closed about their datasets). I am for openness. But also beginning to realize that in some situations you should be careful about sharing your research and results, especially when you’re not certain if the data is correct. Journalists have the responsibility to get the facts right, this is the same for researchers in academia. You are accountable for what you tell. I will try very hard to reach the same standard.

I have made some changes to the analysis framework. I have merged the presentation category with the technology category, so there now four categories: General Information, Presentation & Technology, Data & Participation and Interactive Narrativity.

I was having trouble with two dimensions, but it’s (almost) sorted out. Hoping to get some feedback about it next week.

The last days were mainly about Excel. I just kept thinking that Hal Varian had it completely wrong. There is nothing sexy about sitting behind your computer filling and checking rows and rows of data. How cool that dataset may be.

This thought changed Yesterday. I scheduled some time to continue to read in the Excel book. The first chapters were useful but not really new, but now I am getting into the advanced stuff. And things are becoming quite exciting, I learned Yesterday about VLOOKUP and it rocks! I was already using the table function (with filtering) but this is even more fun.

The last time I was so excited about something similar was when someone explained me how to combine CSV with ActionScript. It was awesome, and that night I explained in detail to a friend how I could use this to parse an excel matrix and create a powerful program. He just looked at me and said: “Don’t get me wrong, but this is not very interesting for me. Lets talk about something else.” And then we talked about other things.

That’s how it is, isn’t? I mean, most of my friends generally don’t have a clue what I am doing. Most saw the graphics overview a couple months ago and said that it was nice. That’s it.

So, I was a little surprised about the reactions I have had the last months about it. It is just an overview. I guess everyone has their own interests and it is probably important to find the people where you can share your interests with.

Conclusion for this update: I think it’s going all right.

I am in the second phase of the data refining. And looking forward to next phase, and secretly already doing some small things, but first need to finish the current phase. I am receiving some more feedback from different people and that is extremely helpful. I am amazed by how generous people are with their time and hope I won’t disappoint.

First, I have been talking with some people within the field. Which is really great. This has led to some adjustments within my academic chapters and to a extensive revision of the analysis framework. A summary of the framework can be found here. Most dimensions are fixed, but two need some more thought. I have arranged to speak next week with my direct supervisor and some other people who are willing to help me.

I have added all new graphics in the excel worksheet and coded the graphics on several dimensions. The first round of coding is done, now it is time for the second round. In the second round I will check the inserted variables and add the new aspects. Afterwards there will be a third round where other people will code random chosen graphics, probably around 20 graphics, to verify if I have been consistent and didn’t make any errors.

In the meantime I am learning some new things about Excel. I am reading an Excel book [1] in linear fashion and begin to use some useful functions in the excel sheet. I don’t want to sound like an Excel geek, but you’re allowed to actually do some pretty cool stuff.

As for the rest, I am trying to keep up with new material. Getting a feel of the current dataset and thinking about what I actually can do with the data when it’s done. And I’m attending different conferences and seminars about visualization (but that is not strictly thesis related, it is mostly out of personal interest).

General speaking, things seem to go well. Don’t get me wrong: there is still a lot to do. But if I can keep up with current speed and nothing unexpected happens, the thesis will be done at the end of March. It means that I have less time for other things – like ice skating, Amsterdam is beautiful at the moment – but that is a sacrifice that I am willing to make.

[1] I am reading “Excel 2010: The Missing Manual” by Matthew MacDonald. It is a great book for beginners, lots of funny examples and everything seems to be pretty well covered.