Web-based Research

Just finished up a small research program that I spent the last half of a year on with the help of a student research assistant. We published our results on the web in a manner that is what I was thinking I wanted to have done over at the Wikiversity, before it got taken over by a hoard of people determine to completely mirror the (broken) university system on the Internet.

I wanted to have a place for scholarship. A place where persons come together to think - write - create - discuss, or as I sometimes put it, for teaching - learning - discusing - creating - thinking. What has happened is that some zealous person or group of persons has gone around and "created" schools and departments for every conceivable area - except that most are stubs with nothing in them. The "original research" page leads in with a "quote" attributed to Benjamin Franklin. Except that this is one of those typical "A said B" things, but no one knows the exact source. I find this also listed as a Chinese saying, and I am unaware of Ben Franklin having been Chinese...

The whole point of the Internet is that we don't have to be in departments, which is what you do when you group things and can only group one thing in one place. On the Internet you can do research that transcends departmentalization. You can link to it from many places. You can publish your data and let others derive alternative conclusions from it, and these can be published there as well.

And this is exactly what we did. We had started out using a Wiki as the basis for our work, on account of me and my assistant normally not occupying the same space-time area. We used the school's wikifarm, which is unfortunately Atlassian Confluence. This is a system which lets me happily write for an hour and then, upon attempting to store the page, announces that it does not feel well and just drops the page (and no, the back button does not work, as they initialize all fields on entry to the page). No, it is not on my list of favorite tools. It would also store pages with umlauts in the name, but refuse to retrieve them. The syntax was primitive and its tendency to really screw up tables just exasperated us to the point that we quit using it.

Since there were only two of us, we just Skyped, sending files back and forth, and set up a WordPress server, which we also used for publishing the findings. WordPress has it's own quirks and needs some interface icing to be really good for this, but in principle it is a great system and did not interfere too much with our work (except for creating tables, which is an exercise in frustration, we ended up making them in Dreamweaver and doing copy & paste on the code).

The research itself ended up being recorded on paper. We had this schema we set up and printed out, but we ended up using the backs of the pages as we discovered that what we wanted to record wouldn't work, but there was lots of other interesting stuff around. We just produced a big pile of paper, and then transcribed the lot. Luckily, my assistant did most of the writing, his handwriting is 10 times better than mine is.

Having it on paper was really good, because we could sort and resort the mess on the floor, until it made sense.

We put together a report, wrote a page on our test methodology and our analysis criteria, linked up all the tests in a big table (including links to the test cases and anything else around), and then we informed the companies we had been analyzing where their report was, and we offered to include a rejoinder, if they cared to write one.

Four actually did, with one going to the trouble of re-analyzing our data for their product (not with the others, though) in order to show that they would come in second. Of course, they would need to have looked at the others and probably given them extra points, too. Whatever - they wrote a four-page rejoinder, which I published and linked from my page.

Of course, understanding the data is not trivial. This is a very complex area. But I have a good feeling that we have our findings, our data, and room for alternate findings, all in the same place. It's not a wiki - so there are no trolls around defacing it. They send me emails with the usual complaints of me "wasting" taxpayer's money on this research. If I'm in a good mood I point them to the constitution of Germany, which ensures freedom of research and of teaching. Of course, if I don't have money I can't do much research. But this was relatively cheap research. And now we can sit down and start to write traditional papers about what we do, linking to the online research collection, which can be updated as I get more data in, I can even include more tests if I so desire!

So I am kind of lukewarm on the topic of Wiki-based research. I would have liked to have had a system that would let us record our results smoothly and reorder them as necessary until we had our findings. And then I need an archive. We piled all the bits of paper on top of each other, printed out the online stuff (I am paranoid) and made a lot of CD backups. A lot of this stuff should not be online, we took a ton of screenshots, only using a few. I like to keep them, but that would just be clutter if I posted them all, and part of research is about inducing order into chaos.

Just some thoughts, maybe this will turn into a paper sometime. Now, gotta get some papers graded.....