RESEARCH PROCESS OF WRITINGCLOUD

2009-06-11T09:39:00.006+01:00

Disclaimer: this post is the diary of the research I carried during the my design process. The experiments were carried from May to June 2009 in order to see how crowds could create material that could be then used later to build a larger body of work. The text created are now printed into the WritingCloudPost, which is a newspaper I printed for the Royal College of Art graduation show.

The post will be a bit long, and will only cover the research. I will post the finished texts in another post.

Here we go

To start this project I needed couple of questions. I asked the closest (geographically speaking) person to me: Ross. Ross is a Mechanical Engineer from Nottingham and really good at asking clever questions that even himself can't really start to solve. So I guess he was the man.

His questions are:

1) Who is the best english football manager in History ?

2) WHich is the most important battle of the 2nd world war ?

3) Where is buried Boudicca ?

4) What does Stendec really means ?

ok, let's go.

gui

I will be using for asking those questions

1) answers.yahoo.com/

2) http://www.answers.com/bb/

3) http://www.convinceme.net/

4) http://www.onlinedebate.net/

5) http://debatewise.com/

I will try to keep these colors all the way long of the writing process.

maybe I ll add some later, depending of the result.

////

I just thought it would be good also to ask them:

Is crowd sourcing to a certain extend exploitation of the masses ?

I just posted the questions on these websites

It took me about 2 hours, registration included. A lot of them were long and boring ( especially yahoo answer) and the learning curve quite bad for most

Answer.com is actually not allowing a crowd to answer so there is no point.

Answers came nearly instantly on yahoo answer, with 8 answers and 5 votes about world war 2 after 1h30, not bad.

I am very impressed also by the quality of these answers. Most of them are properly written and argue their point clearly. Nevertheless no discussions seems to start between the people answering to build a consensus. Everyone is just coming and adding his point to the list. It feels a bit like a show of knowledge rather than a will to build.

On Onlinedebate, which is more of a small-ish forum, you can feel that the community is at home, it is a pretty different atmosphere than the others which are definitely more anonymous. I don't know yet if the one to one kind of interaction is pleasant and relevant or not for my purpose. Also the community aspect allow more freedom to experienced members, which is either good or bad.

edit - i ve been just asked (30 min or so after posting) to edit my questions and add more material to it, or they will be removed from the forum

Debatewise is still in beta phase, therefore not really handy to maneuver. It is impossible to have a direct link to the debates you created. it is a bit frustrating.

The reaction from the onlinedebate community have been very quick, polite and firm. In an hour my questions have been deleted and a bunch of people visited my profile to see who I was.

A user reminded me of the forum rules in my personal message box and the administrator told me in a very nice way that he will remove my questions but i was more than welcome to post them again if they were to a higher standard.

This is a two fold learning

1) it is an AMAZING example of self regulation. I suspect that my account would have been deleted if I had show any sign of rebellion

2) Like in many forums experience is a key to acceptance. The structure is very pyramidal and you will have to prove your good will before being fully accepted.

so are still on the way

1) answers.yahoo.com/

3) http://www.convinceme.net/

5) http://debatewise.com/

the experiment is now running for 3 hours and the results are IMPRESSIVE. A person wrote a full lenght article on conviceme, arguing that D-Day is definitly the most important battle in WW2.... stunning

Also a young lady called eireaislinn on yahoo answer is very opiniated about Boudicca. She apparently wrote an accademic paper about it. I will try to contact her to see if she can help me more about this subject

I opened a wiki to aggregate the results at this URL

http://www1.atwiki.com/writingcloud

It was very very easy and took me less than 5 minutes for a "key in hand" solution.

I do remark after nearly 4h of work that:

1) it is very tedious to refresh manually all the pages where you asked your questions, in my case about 15 in 3 different website

2) I ve been working on the wiki and body of the articles while waiting for answers

3) Editing text in the wiki i use is very bad, and i suspect it to be bad in any wiki.

4) the copy/past thing is pretty natural

5) I still manage to update this blog live at the same time whithout feeling overwhelmed

6) The "My question" box from yahoo answer is very handy (image at top)

It is 9 o'clock and the experiment is now running for nearly 12 hours

I went to sleep yesterday at 1 o clock in the morning. I discover couple of things this morning

1) all my questions aren't actually deleted from online debate and people are answering them. SO debateonline is back in the race

here is the list of the website I am using

1) answers.yahoo.com/

3) http://www.convinceme.net/

4) http://www.onlinedebate.net/

5) http://debatewise.com/

2) the evolution overnight is mitigated

On onlinedebate there is currently

>> 4 answers about boudicca. All of them are kind of jokes, ranging from free mason conspiracy to things with chuck Norris. The forum seems mainly American, this maybe why this very English topic does stick well. Also, as the first reply was a joke, the community members apparently fell obliged to keep the joke going.

>> 8 answers about the best football manager, some names are given but with very few reasons why. The discussion is mainy lead by two english people and some americans are doing jokes about british and soccer.

>> 5 answers about crowd sourcing: ranging from simply " no because it is the choice of the people" to "what is this word that i don't know"

>> 2 answers about WW2, mainly taking D day as the most important battle with couple of explanations why.

On yahoo answer

the overall quality is very good, succint and precise. Yet no evolution overnight as I guess the questions drawned in the abyss of the website very quickly

>> crowdsourcing: 0 answers

>> Stendec: 0 answers

>> Boudicca: 5 answer / 0 overnight

>> WW2: 7 answers / 0 overnight

>> Best football manager >> 2 answers, / 0 overnight

On convinceme

>> WW2: 4 answers / 2 overnight

>> Stendec: 0 answer

>> Boudicca: 0 answer

>> Football manager: 2 answers / 1 overnight

>> Corwdsourcing: 1 answer

Most of the answer are no longer than a phrase, only the WW2 question attracted a huge answer, nearly an article in fact

On debate wise

no sign of life whatsoever...

Conclusion

So far i consider Yahoo answer giving the best result for my purpose. The number of people answering is quite important, as long as subject is appealing. THis is maybe why the football manager question didn't work. this cross correlate to a cetain extend the Technorati study saying that bloggers were watching a lot less TV than the average person. Also he question is very specific.

Nevertheless it seems that touching the right people with the right questions is crucial to receive a quick and documented answer ( convinceme). In that case this happent by luck. If this could be done cleverly I am sure this would raise the level of the answers.

Regarding onlinedebate, the community aspect seems quite a barrier to my purpose as the penetration rate for newcomers is very weak and the existing users are there to be with each others rather than really answer questions.

Also, the "deleting question" reaction from the administrators, seems to be more of a defensive reaction against someone too expansive and looking like a thread of some sort, rather than a real care for quality of thoughts of the forum. Further researches shown that most of the questions/topic are very very weak in their arguments and potential for debate and my questions started discussion without needing any improvement.

The discursive aspect of a strong community is also most of the time leading to move from the initial subject to trivial remarks about the wife, the dog, the car or again jokes of some sorts.

Again, targeting the right people is what seems essential.

I decided to leave this experiment run for 12 more hours, thus ending around 8 tonight after 24 hours.

For the next ones I would like to

1) write about a "current affair" rather than history or personal opinion about a obscure subject.

2) go deeper in one of the current subject, by asking more specific questions. I think so fare the WW2 topic seems the most appropriate

I finally had the experiment running for 36h and here are the results

I asked questions on four websites, three of which were responsive

1) answers.yahoo.com/

3) http://www.convinceme.net/

4) http://www.onlinedebate.net/

On the five questions i asked one stayed totally unheard (what does stendec means ?) and another one (which is the most important battle of second world war ?) created real debates in all the responsive sites.

It became hard and irrelevant to count the number of answers after a while as a lot of them are either trivial, jokes or aknwoledging quotes from previous post. The discursive aspect was in a way beneficial as it kept the thread in the top positions in the forums, but mainly the biggest benefit was to oblige some contributors to build stronger arguments and, sometime, bring sources

Here are the amount of words I collected for each topics after these 36 hours

WW2: 2354 words

Crowdsourcing: 393 words

Bouddica: 196 words

Manager: 194 words

Stendec: 0 words

Total: 3137 words

I will now start editing the text to build articles I ll post on this blog. I guess the questions will still evolve in the meanwhile but for the purpose of this experiment I have the feeling that 36 hours is enought.

I also deceided, as the world war 2 topic seems very responsive to dig it with the same websites and in particular Yahoo answer. Because the others are forums of a smaller size I will need to be weary to not be seen as a spammer or trouble maker.

Interestingly this might shows that anonymous and uninteresting way of asking questions on the internet isn't a good pay off for my purpose

I thought it would be interresting to give here some data about the lenght of the average blog post.

This may help to compare the amount of data I collected to the average blogosphere.

According to the following figures I am not that badly served as i collected more than 10 times the lenght of the average blog post for the world war 2 questions.

The other questions are in the range of an average post lenght

source http://modernl.com/article/how-long-is-the-ideal-blog-post

Taking the Technorati Top 100 as my sample, eliminating those which aren't in the English language, and those which aren't identifiable as conventional blogs, I took an average word count of the 10 most recent posts on each blog.

The graph above shows the distribution of the average word count - the shortest average was a terse 77 words, and the longest a snooze-inducing 1,449. The vast majority of posts were over 100 words and under 500.

If we group the posts into distinct strata, we can see a clearer indication of the distribution:

Essentially, the distribution is a normal curve with positive skew - with half of all blogs having an average of 100-249 words per post.

Consistently shorter posts are rare - usually limited to blogs devoted to non-written content (pictures, videos, audio etc.), and with celebrity gossip blogs having a tendency to be more concise -Perez Hilton, for example.

On the other end of the spectrum, there's a distinct jump in word length for some blogs - in these cases it's usually a transition between writing for entertainment to writing for reference. With the transition comes the ability to post much longer articles, without the fear of losing interest. The articles over at A List Apart are a good example of this.

I just posted the last article. It took me about 2 hours in total to edit the 4 articles from the bits and pieces I collected on the 3 websites. I spent most of my time editing the one about second world war as it is the longest (about 3000 words). Nevertheless I mainly worked on articulating paragraph and didn t touch the body of the text that much. Strangely enough, people who were writing on different forums complement each others really well.

This may introduce the fact that multiplying the sources is beneficial to the overall body of work. just like writing a paper in a normal way.

It is a real sucess for a first trial I think as I spent maybe in total 5 hours working on these questions ( registration on all the website and learning how to use them included) and I managed to write 4 relatively OK articles.

In total they are more than 4000 words, so I "produced" around 1000 words an hour. Not bad.

Now i will focus on increasing the over quality of thoughts and the referencing. I m planning tomorrow to find forums specialized in certain subjects and ask a very precise question, sources included.

This will help, I hope, to prove that reaching the right people when asking a question on the internet is crucial to receive in return a high quality content and experience a high reactivity from the crowd.

It has been very exciting so far as most of my assumptions seem to be verified. Looking forward to the next round.

It has been about two weeks I didn’t do any other experiments due to work overload but I managed the last two days to do a new one which is pretty interesting.

Also I did film all my activities during the experiment so I will be able to post a video soon.  I bought the Financial Times from 1^st of June and grabbed couple of articles. From the body of those articles I extracted material to build questions. These questions were a bit more documented than the ones from the former experiment.

Here they are:  

 Will MP’s expenses scandal influence the coming European elections?

The scandal continue to cause reactions as Alistair Darling, in charge of the UK’s finances seems to have claimed for six month the rent for a flat in south London while also claiming allowances for his free apartment in Downing street. With the European election coming, do you think this scandal will influence voters?

Cattle’s farming is destroying the rain forest in Brazil. How could we avoid that without damaging the Brazilian economy?

Cattle farming account for more than 7 billion dollar a year in Brazil’s exports. However, this practice is the most important cause of the tearing down of the rain forest. How could it be possible to solve this issue?

 Is North Korea about to start World War 3 or is it only building tensions to attract attention toward itself?  Satellites spotted North Korean Long distance missile being transported by train this week. North Korea got a history of shifting equipment to attract attention and generate crisis, knowing that US spy satellites would detect those movement. However, North Korea fired last month a long-range missile over Japan. Where do you think this lead?

 What would be the best strategy to end the war in a responsible way in Afganistan ?

Last month the Obama administration announced its new strategy in Afghanistan and Pakistan. A few days ago, Robert Gates, the defence secretary, said the US and its allies had a year in with to make meaningful progress. What these progresses should be?

Could teaching business be the solution to the 3^rd world poverty issues ?

The French HEC management school is currently helping producers in Burkina Faso to become more business minded. If this measure was globalized, could it be the way for solving world’s poverty problems?  

Instead of posting those questions in random/generic places like yahoo answer I looked after forums with focus into the questions. I failed to found a very relevant forum for the Brazilian cattle and teaching business in Africa.

Here are the forums/website I used from last time:

Yahoo answer convinceme.net Onlinedebate.com  And the new ones:  Politicalcrossfire.com UKdebate.co.uk debatepolitics.com Politicsforum.co.uk

I was blocked by moderators on forum.greenpeace.org

The reasons for this blocking are unknown on the Greenpeace forum. I suspect that they just dislike political subject and want to keep a very strict “editorial” line that is directed toward the image they re promoting.  Anyway, it took me hell of a lot of time to navigate, register and understand how all these different forums worked, as they are all different from each other’s. Also navigating in the maze of categories and subcategories was very confusing. It took me about 2 hours post these five questions on theses 8 platforms.

I left the experiment running for 36 hours. From midnight the 1^st of June to midday the 3 rd of June and the results were amazing. I did gather 8 129 words of relevant data (I didn’t copy the one line comments and jokes) and a lot of link to very relevant documents and videos.

Afganistan >> 2732 words 

North Korea >> 4099 words 

MP’s expenses >> 653 words 

Brazilian cattle >> 228 words 

Business in Africa >> 417 words

Learning outcomes

1) The crow is extremely receptive to news related questions. The North Korea subject was just amazing, with more than 4000 words in 24 hours. This is great news for writingcloud as a way to engage with breaking news.

2) I learned from the first experiment that organisation of bookmark and text is key. Still this methodology is extremely time consuming as it took me nearly two hours ( from 12 to 2 o clock on the 3^rd of June) to harvest the text posted by the crowd. This is just very inefficient and brings more validation to writingcloud.

3) When questions become more specific generic platforms yahoo answer become pretty useless. The age of the audience might also be a reason why more political questions didn’t work well in yahoo answer.

4) Finding the right people that will engage with the subject is very important. For instance, The MP’s expenses attracted very few reactions from American forum and very few people were interested in the cattle farming in Brazil

I will now edit the texts and post them to the blog.

Nice arty things

2009-06-09T15:16:00.005+01:00

I thought i d share with you some nice art project a came across during my research

1 - http://www.thesheepmarket.com/

Artist Aaron Koblin's phalanx of 10,000 sheep, all drawn by random strangers through Amazon's labor distribution mechanism, The Mechanical Turk. Koblin's goal was to raise questions about the emergence of new labor systems in the information age.

2 - christopherbaker.net

Murmur Study is a work-in-progress that examines the rise of micro-messaging technologies such as Twitter and Facebook Status Updates. One might describe these messages as a kind of digital small talk. But unlike water-cooler conversations, these fleeting thoughts are accumulated, archived and indexed digitally by corporations. While the future of these archives are still to be seen, the sheer volume of publicly accessible personal expression should give us pause.

These first studies presented at the 2009 Spark Festival and the Experimental and Media Arts Exhibitionconsider the absurdity of the physical archive by printing live Twitter message from the Minneapolis Metro area during the exhibition. Twenty thermal receipt printers continuously produce and endless waterfall of text, which accumulates in tangled piles at the bottom.

Murmur study is an ongoing collaboration with Márton András Juhász and the Kitchen Budapest.

more to come, stay tuned

WritingCloud

WritingCloudPost

RESEARCH PROCESS OF WRITINGCLOUD

Nice arty things

Facebook group

Hello world :-)