Saturday, November 15, 2008

Thinkers, Scientists or Mechanics

Kia ora tātou - Hello everyone
A map of brain activity according to TypealyzerMy brain activity while I write this post - according to Typealyzer.

Bloggers are generally categorised as thinkers, scientists or mechanics. That’s the conclusion I might have come up with when I decided to find out more about how Typealyzer categorises blog sites listed in my RSS Reader.

Michele Martin, Stephen Downes, Amy Gahran, and others wrote posts on the latest Web2.0 download for performing a Myers-Briggs analysis from a blog address. Stephen declared (I think with tongue in cheek), “The fact that my site turns out to be INTJ - exactly as I test consistently on the Myers-Briggs tests - is coincidence. Right?”

Data validity:

The inference is that data from a blog site can be used to determine the personality of the writer. Hmmm. This could have implications to do with data validity, if a blog site was used to categorise the person who writes posts on it.

I decided to do a bit more exploring. What I found gave me food for thought. By the way, the Typealyzer put my blog site in “The Thinkers” category. The picture above is supposed to be my brain activity while I write this post.

Anyone who has visited my blog and inspected my blog roll will be familiar with some of the bloggers whose names I list here. Yes. You’ve probably already worked out what I did.

Site analysis:

I went through the list of blog sites in my RSS Reader. I took each site and pasted its address into Typealyzer. It didn’t take long. A few sites incurred errors for one reason or another and couldn’t yield any data. There were other sites I visit less often that I left out of this study.

Apart from Mathew Needleman, whose site yielded “ISTJ The Duty Fulfillers”, here are the sites listed according to the names of the bloggers associated with them. I have deliberately listed their names in some sort of random order.

A table of blogger's names arranged according to Typealyzer analysisAdmittedly, this analysis is entirely anecdotal. The sites were certainly not chosen at random for they are all selected from my RSS Reader.

What could it all mean?

I mused over the results. The most prevalent sites were the INTJ and INTP types. Of all the least likely personality types I’d meet in a crowd of 100 people, the INTJ and INTP would be among them (according to Myers-Briggs USA inferential statistics). In fact, together they would represent just over 5%. That’s 5 people, in a room of a random sample of 100!

I also wondered about the sheer lack of IS types other than the mechanics who seemed to be almost as prevalent as those in the two other groups. If they'd followed statistics, however, the mechanics should have been more numerous.

My selection?

I recalled Michele Martin’s posts on homophily and wondered if it was something to do with my choice of blog sites. Have I been unconsciously selecting blog sites according to their 'personality'?

Or is Typealyzer not really doing a valid analysis? Stephen Downes’ and Michele Martin’s evidence, in support of Typealyzer’s accuracy, seems compellingly convincing. Using USA inferential statistics, their claimed personality types are among only 2.1% of the population. Yet the group they are in is the largest group listed in my table. There are 13 other group types not listed here because I simply didn't find any more. Where are they all?

All sites in my blog roll are written by bloggers who have an interest in training or education. Could this be a factor, an artefact of how they are selected, that has put them into these exclusive groups?

One last thought. I was introduced to Amy Gahran’s site through Stephen Downes’ post. She found that her site returned “ENTP The Visionaries”. When I entered her site address on Typealyzer, the return was “ESTJ The Guardians”.

I wonder if I was holding my mouth right.

( 2 ) << - related posts

Haere rā - Farewell


Sue Waters said...

Well I am a scientist so it's good to know I fit into that category if the results are correct?

Although face to face (apparently) my nature can be frustrating because I tend not to take anything on face value and spend considerable time debating the matter. Not to mention hours research the topic. (Drives my husband crazy)

Blogger In Middle-earth said...

Kia ora Sue.

You say you "tend not to take anything on face value and spend considerable time debating the matter." And the research. Sounds as if you have Carl Sagan's Baloney Detector working down to a fine art.

Over 20 years ago I was Myers Briggsed and found to be INTJ - The Scientist, as you are. I had confirmation of this over the years and worked hard at trying to balance some of the finicky quirkiness of the scientific approach out of my life.

I don't think I've succeeded in shifting much! I'm now an INTP according to Typealyzer, if it's to be believed, which means I sit around and think about it all now :-)

ka kite

murcha said...

I wonder where I fit in to this and I am not game to find out. Just visiting to thank you for joining in the orange day in a sentence stars.

Michele Martin said...

Very interesting results, Ken and like you, I'm not sure where they're coming from. It would make sense that you'd be attracted to people who tend to think/write in similar ways, so maybe this is another example of homophily playing itself out. If that's the case, maybe we can get more diversity into our readers by first running URLs through the Typealyzer. :-)

Britt Watwood said...

The scientist in me says that this is fascinating! It also does not surprise me that you were classified as a is the same way I think of you! Nice job!

Ken Stewart said...

Ken, this is interesting. I am always curious about things such as this. Although the results are somewhat consistent, I always question the means in which the judgement was made.

So, in effect, I suppose I am confirming I am a scientist ;-)

It is also funny in that I have 2 sales-types who write in my feed as well... I wonder if it considered those in the equation?

Blogger In Middle-earth said...

Tēnā koutou katoa!

Though I didn't explain this in my post, the Myers-Briggs (MB) categories, Scientist, Thinker etc, are only convenient labels. Someone with the personality 'scientist' may not necessarily know the first thing about Science per se.

The fact that a scientist like you or I, Sue, happens to have pesonalities (according to MB) serves only to indicate that our categorised personalities are suited to scientific study. That I have a PhD in a scientific discipline AND the personality to go with it probably means that I am suited to a particular approach to Science, which is probably why I got my PhD in the first place. Frankly, my preferences have always been more towards the Arts.

The same can be said for you, Ken, though I have no idea what your background in Science is. I presume that you have a business background (a hollow assumption, I know). But a 'scientist', as such, can have any of the MB personalities and still be an effective scientist.

Homophily is one thing Michele but this sort of exclusive selection is another. There are 13 other categories, according to MB, and I've apparently missed them. Perhaps I should now be on the look out for those bloggers (if they exist).

Britt, thinker or scientist, I don't feel any different from when I first started blogging. Thanks for the confidence that, at least, I appear to be doing a bit of thinking. I think I need all the thinking skills I can lay my hands on at the moment.

Haere mai Murcha. I am delighted that you have dropped by. If you don't already know these people, may I introduce you to Sue Waters, a scientist who lectures in aquaculture and somewhat of an icon in blogging, Michele Martin, a freelance consultant and also an icon in blogging - one who has put me on track-so-to-speak.

Britt Watwood is an online learning specialist at Virginia Commonwealth University’s Center for Teaching Excellence, and fellow thinker and blogger, who has recently confirmed his blogging personality as being aligned with the scientific discipline.

And Ken Stewart, who is, I believe, a leader in the technology business, another blogging companion and who, like Britt, has been introduced to the idea that his blogging style is aligned towards a scientific bent.

Haere rā

Mattias Östmar said...

Hey everyone!
Thank you all for your feedback! First of all, let me emphasize that is NOT the same as a proper MBTI-test for god and for bad. Typealyzer is a way for us to test our hypotheses that it is possible to analyze the persona of a person by teaching a classifier to identify word usage among different personality types. An individual can (and normally DO) make use of different personas, i.e. social roles in different situations.

Our hypotheses is that an individual often makes consistent use of one persona on a blog, but might have different blogs with different personas - such as a family-oriented and a professionally oriented. And professional writers/communicators are most likely using different personas in different texts purposely. is a free web service that helps us develop our technology through user feedback and hopefully at the same time is a fun way for bloggers to get interested in personality types.

There is so much cool research to be done! Blogger In Middle-earth points out one of the things that made me start this project nearly two years ago - namely that there seems like there might be interesting patterns in what blogs we choose to read and engage in depending on type preferences.

I would like to perform large scale studies on linking behaviour to see if it might be so that for instance thinkers tend to read and engage in thinkers blogs and idealists in idealist-blogs etc. I think so.

Blogger In Middle-earth said...

Haere mai Mattias!

I am honoured that you have visited and given us some insight into the history of Typealyser, and how it works.

Obviously, as you can see, there’s been a bit of speculation here about its function. I realise it’s still early days yet, and there will be refinements made to the design – is this right?

If the Typealizer is giving us valid analysis of the text contained in the blog sites, and let’s assume it is, does this mean that the results that I have gathered recently and displayed on this post indicate my preference for a particular type of persona, as you call it, in blogs?

I’m sure that you understand that only three major groups of blogs, categorised according to Typealizer analysis, leaves a huge number of blogs (perhaps even educational blogs) that I’ve either rejected by one means or another or simply haven’t come across.

Given that many of the people listed on my post know of each other on the blogosphere, does this mean that educational blogs tend to fall into certain exclusive Typealizer categorisable groups?

If this is so, and it can be proved or validated in some way, this would make very interesting information from the point of view of working with online communities, especially those groups that work with such web2.0s as Ning etc.

I would dearly love to be involved in your study. The wider picture for me is to gain an understanding of what engages learners in discussion, from the point of view of an educator, rather than for categorising blog writing types - thopugh I realise that there would be a close relationship.

I am sure we have a huge number of educators who are watching this conversation right now, and who would also be as interested as I am in finding out more.

I’d really appreciate your follow up on this

Catchya later

Laura ~Peach~ said...

Kia ora :) I had to come check out your blog after reading your comment on fever bee... I was learning some maori a while back and unfortunately dont recall much now but it made me grin and brought back wonderful memories of an internet friend from Auckland :) thank you for the rush

Anonymous said...

Hi again blogger in middle-earth!

Thanks for great questions and feedback! As you say we´re definately in the early beginning of this work.

It was very recently that we started to segment/divide the texts into 16 categories according to C.G. Jung and Isabel Myers-Briggs and David Keirsey´s later work.
If you run blogs through the Typealyzer today it should give you these 16 types instead of the four categories that we used to present.

Hopefully this will make it possible start investigating networking behaviour not only between blogs, but also in communities and, really, every other linking and meme-spreading behaviour online.

To be able to this we need to develop further design. We actuallt have been spending most of our time developing a reserach tool we´re it´s possible to input large amounts of data (such as blog posts mentioning a certain thing or linking to a certain site or whatever), analyze it and present the type data in neat diagrams. See some first examples on the video on my blogs first page:

Right now we´re thinking about how to get help from people like you to do both qualified studies done and help us provide further tools for other interested people at universities, companies and individuals. We´re thinking about giving access to our classifer through an API - would that be helpful to you (needs some programming skills)?

It´s part of my vision for this that there will be lots of good applications of this - both for fun, learning and commercial purposes.

Blogger In Middle-earth said...

Kia ora Mattias!

Thank you for the link and your generous information. As you may have already seen, I've posted your video.

I will be fascinated to watch how your research develops, and would be grateful to be involved with it in any way, if at all possible. Good luck with that.

Ka kite

Sue Waters said...

Hi Ken, Vicki Davis just used Typeanlyzer on my blog and now it gives a different result. I was wondering if you could drop past her post to see if they are now including more categories and that is the reason why. The scientist in me is wondering the why?

Blogger In Middle-earth said...

Kia ora Sue

I have an explanation, though you may also have to do some scientific investigating here ;-)

I have a suspicion that I did not analyse the Edublogger site but that the result I posted was for Mobile Technology in TAFE. Since the Typealyzer really analyses the persona of the blog and NOT the personality of the blogger, it is concievable from a valid scientific point of view that the results could be different.

I am unable to check, as my PC is gone bung right now, but my memory says that this could be the reason.

My blog vs me

As an aside observation: it's interesting that for as much as I have attempted to trumpet the difference between blogger personality and blog persona on all subsequent posts and comments I've made since Mattias first commented here, people still believe that the Typealyzer analyses the personality of the blogger. It doesn't.

By the way, I am an INTJ - which some authorities would say is The Scientist (nothing to do with Typealyzer).

Typealyzer analysed my blog as INTP - The Thinker. Chances are my personality is a bit of both, but my blog tends to express more of my thinking than my scientific proclivity.

Catchya later

Blogger In Middle-earth said...

Kia ora Sue!

I've done the homework for you. Aren't I good?

In fact the most recent results show:
Edublogger - ESTJ - The Guardians,
Mobile Technology in TAFE - ISTJ - The Duty Fullfillers.

Neither of those agree with the result I got in November 2008 and posted here.

So, unless Mattias can afford some explanation, I'm stumped, other than to suggest that, perhaps, your blog(s) have changed their persona significantly by your recent post contributions.

Let's see if Mattias picks up this conversation.


Sue Waters said...

No. I did check it on both blogs and was duty filler for both. Rechecked again and is the same. Weird since you are getting different results from me?

PS if you used WordPress powered blog you would have his email address :)

Blogger In Middle-earth said...

Kia ora Sue

The difference is there it is true, but frankly it may well be according to a number of different things. As it happens, I observe that the STJ part was returned for both when I did them recently which is at least in 75% agreement with what you got. Averaged out you could say my observations were in 87.5% agreement.

My summary of the whole thing is that it is a bit of fun, which is what Amy Gahran said, and what I catalogued this post under in my index.

As for emailing Mattias, perhaps Vicki Davis could do with some of your advice on this. But I think it will be an interesting exercise in itself to see if Mattias gets back to this conversation ;-)

At the moment I'm happy to have a blog even if I don't have a PC.

Catchya later

Blogger In Middle-earth said...

Haere mai Laura~Peach~!

A belated welcome to my blog! I think I dropped by your blog - colouful and interesting - and got distracted! Hence I forgot to welcome you here!

Do drop by again some time.

Catchya later