What exactly Is a Theory?

I’m on sabbatical at the moment and so I feel entitled to think Big Thoughts. However this one is a rather embarrassing thought. The more I think it the more I realize that I’m not really very sure what the word ‘Theory’ means. Maybe it doesn’t matter. Maybe ‘theory’ is just one of those Humpty Dumpty words that means whatever the speaker chooses it to mean. But I’m often present in situations including academic job interviews where person A asks person B: “So what do you see as your theoretical contribution?”  I worry because I’ve no idea how I would answer, given that I’m not at all sure what a theoretical contribution is, let alone if I have one or more to make.

My background is in computer science and I still consider myself to be a computer scientist, albeit a rather odd one. I now work in a School of Library and Information Science that is also an iSchool. CS doesn’t seem to be bothered too much by theory. A lot of it is about building stuff that is better-faster-cheaper. Or that is more powerful or generic. Or building processes or things that allow other people to build things better-faster-cheaper. Yes there are theoretical computer scientists and I think I know what they do – looking at the big picture abstractions of invariants across all or whole categories of programs or computation or whatever. But they are definitely allocatable to a special  category, and CS departments normally have just a few of them kept together in out of the way places so as not to frighten the students. It is rumoured that they use blackboards and consort with mathematicians. They show what they do is right, not by measuring it against other people’s stuff and showing theirs is better, but by doing proofs.

It is very different in LIS. Everyone seems to bang on about theory and to expect everyone else to do so too. But what do they mean? Do they all mean the same thing? Or are there at least some aggregations of agreements?

I used to think I knew but now I’m having doubts.

Some definitions

A scheme or system of ideas or statements held as an explanation or account of a group of facts or phenomena; a hypothesis that has been confirmed or established by observation or experiment, and is propounded or accepted as
accounting for the known facts; a statement of what are held to be the general laws, principles, or causes of something known or observed. (OED) cribbed by me from (McGrath 2002)

And from (McKechnie & Pettigrew 2002):

• “A set of explanatory concepts” (Silverman, 1993, p. 1)
• “a statement or group of statements about how some part of the world works–frequently explaining relationships among phenomena” (Vogt, 1993, p. 232);
• “a systematic explanation for the observed facts and laws that relate to a particular aspect of life” (Babbie, 1992, p. 55);
• “generalizations which seek to explain relationships among phenomena” (Grover & Glazier, 1986, p. 228);

So those all sound very nice, friendly and manageable. I understand them. Any halfway decent research paper should have things in it that can be described by one or more of the above. So does that mean that all such halfway decent research papers have a theory in them? I’d be happy with this rather universalist definition of theory, but I’ve a feeling that a lot of my peers wouldn’t. I think they mean something stricter and rarer. It’s just that I’m not sure what it is.

Well I’m not going to solve this puzzler today. I can see I’ll have to do more reading, thinking and talking before I get any clearer.

From Analysis to Synthesis

But, in the interests of true academic game play herewith are a set of theories about theories. That obviously makes them metatheories and therefore count double points. I present them as also hypotheses to be tested in my ongoing reading.

• The Declarative Metatheory. A theory is a theory if someone says it is a theory
• The Nominative Metatheory. A theory is a theory if it has “theory” in its title
• The Acclamatory Metatheory. A theory is a theory if someone other than its author cites it.
• The Teleological Metatheory. A theory is a theory if it does something useful in helping understand or predict something other than the case or example it is derived from
• The Obscurantist  Metatheory. A theory is a theory if it uses long complicated words to say something that is blindingly obvious if translated into plain English
• The StrictInterpretationist Metatheories: variants of the above with ‘if’ replaced by ‘iff’

Now we’re on a roll. If we can have theories about theories, why not theories about theories about theories?

So with a roll of drums, ladies and gentlemen let me present:

• The Classificatory Metametatheory. All theories can be described by one or more of the preceding metatheories

References

Babbie, E. (1992). The practice of social research, 6th ed. Belmont, CA: Wadsworth.

Grover, R., & Glazier, J. (1986). A conceptual framework for theory building in library and information science. Library and Information Science Research, 8(3), 227-242.

McGrath, W.E. (2002). Introduction: library science and theory. Library Trends, 50(3), 309-16.

McKechnie, L. M., & Pettigrew, K. E. (2002). Surveying the use of theory in library and information science research: A disciplinary perspective. Library Trends, 50(3), 406-417.

Silverman, D. (1993). Interpreting qualitative data: Methods for analysing talk, text and interaction. London: Sage.

Vogt, W. P. (1993). Dictionary of statistics and methodology: A nontechnical guide for the social sciences. Newbury Park, CA: Sage.

The Usability of CHI

Starting in the 1980s a lot of usability professionals wrote about the challenges of introducing usability into commercial settings. They listed a number of different reasons (or better put, excuses)  they were given why a organisation and particularly software developers and their managers might be resistant to the human computer interaction approach.
Here are some commonly described objections/excuses not to improve the interface:

• It works just fine for me
• We can’t change now it because that would make us look bad that  we hadn’t changed it earlier
• I’d like to change it but we need to get everybody in the entire organisation who uses the product to agree and that’ll take so long it’s probably impossible so let’s not bother
• It’s just a matter of learning how to use the current system. I know it’s a bit challenging but you just have to make the effort
• The current way is very convenient for the people who develop the product and your proposal would make it more difficult for them just to benefit consumers. I don’t think that’s going to fly round here
• Look, we’re a monopoly supplier. Our customers just have to accept what we give them. We don’t have to make it more usable for them
• We’ve always done it this way
• We can’t change it now because then it would be inconsistent with all the older stuff that we’ve produced over the years
• The engineers in the organisation like it. They don’t really care about what customers think.
• Well that’s a really really strange idea – thinking about how people would actually use our product. We don’t really think that way here. We just think about how we can produce stuff most efficiently
• Your proposed ‘improvement’ will take up more space. We can’t be wasting precious resources just making it look nice
• Isn’t what you’re proposing just an issue of aesthetics; just making it look pretty? We are much more interested in the content not how it looks
• Oh, it’s all so subjective.  Just because you say you don’t like it doesn’t mean other people also won’t like it. We deal in cold hard facts and we can’t have any of this opinion nonsense getting in the way of our processes

I’m sure many of you recognise these kinds of objections especially from  people we  stereotype as cold hard engineers who have absolutely no empathy with the needs and struggles of the end-user – for whom of course as usability professionals we are the advocates.
Well, so far so good. Now for the Judo throw. I’ve just been doing some CHI reviewing. If you don’t know, CHI is the main conference for Human Computer Interaction researchers. Getting a paper into CHI is A Big Deal. Indeed in some ways it is more of a deal than getting a paper into a journal, because we’re all too busy nowadays writing things to bother reading what other people are doing.  I want to draw your attention to the usability of CHI itself (and by extension other ACM conferences). After all isn’t a research paper just as much an interface as a webpage or a computer application?

So, what about all these references? We use a very strange bibliographic convention in the ACM publications with  numbered references. It isn’t very consistent and bears only passing resemblance to other conventions. But today I’m not so much looking at the references listed at the end of the paper, but how we refer to those references inside the paper. Why on earth do we use numbers like [6] rather than author-date like (Fu & Pirolli 2007)? Are numbers really the best way to cite work when you’re writing a research paper?  Do numbered references improve the user experience of reading papers or do they rather get in the way when you’re reading a paper? Has anybody even asked this question?

There you are reading and you see a citation of [7]  Oh yes, you think, good old 7. that old chestnut. Yes of course she would be citing him to make her case wouldn’t she? Yes that totally changes your understanding of her argument. There she goes bringing 7 in to back up her mad idea, and worse she dares to try and  demolish [8]‘s claims which you rather liked. She’ll need to give you strong evidence that 8′s wrong after all these years.

No of course that’s not going to work is it? I’m sure you’ve experienced that flopping back and forth to the end of the paper to remind yourself what is 7 and what is 8 even when you’ve read both. Wouldn’t it be far better to follow the APA style convention of including name and date and in our references?
Wouldn’t that improve the user, the reader, experience significantly?

So what are we going to do about it? I wonder what would happen if we proposed this change to the ACM on the grounds of usability and the user/reader experience? I wonder if there would be a flurry of objections. I wonder what the objections would be. I wonder if they would look suspiciously like the list above… It would be a fun experiment wouldn’t it?

Now I have to be fair and balanced. After all, including the name and date all over the place just gobbles up extra space so it will be very inconvenient for the writers of a 10 page conference paper. And let’s be reasonable – maybe the number of authors of the paper actually outnumbers those who are ever actually going to read it carefully enough to pay attention to an argument based in part on citations.  So maybe optimising for the authors of the paper to cram more ideas into their 10 page limit is the right way to go. But somehow I have my doubts. I do hope it doesn’t apply to my own papers.

Finally, let me quote from an overview of APA style:

“The best scientific writing is spare and straightforward. It spotlights the ideas being presented, not the manner of presentation. Manuscript structure, word choice, punctuation, graphics, and references are all chosen to move the idea forward with a minimum of distraction and a maximum of precision.”

Found at http://apastyle.apa.org/about-apa-style.aspx

Doesn’t that sound like a really nice user centred design rationale for their bibliographic convention?

I’ve not been able to find the equivalent rationale for the ACM style. Do let me know if you find one.

CHI and Reviewer Modelling

As the CHI deadline approaches I thought it might be handy to share a heuristic I find useful while writing. Note I’m not particularly successful at getting into CHI, but I have a lot of experience of failure. Bearing that in mind, here’s my tip:

• Try and channel your inner CHI psycho-reviewer

I’m sure we’ve all had reviewers’ comments that make you doubt the sanity of the writer. There are those that are almost willfully obtuse in failing to see what you are talking about. There are the egomaniacs who only read anything in terms of how it relates to their research and rate everything accordingly. And there are the members of the Methodological  Inquisition who sniff out any hints of experimental heresy. With most of these there is little you can do other than to hope you don’t get too many.

But there are some slightly more normal people reviewing, whose reactions can be predicted and where a bit of rewriting and reframing can help in the process of reviewer-proofing. It can help to use the lawsuit metaphor that we already borrow from in talking about a thesis ‘defense’. As the defending counsel in a case, our job is to guess what the prosecution will say in advance and pre-empt it, rather than relying just on a rebuttal.

So what are some classic attacks in reviewing? Here are some. You may have others.

• I don’t see what’s new here. Hasn’t this all been done and said before? What’s the value-add?
• Isn’t this just a reapplication of work X in context Y? I’m prepared to concede you are the first to do it in Y, but is it really that different?
• This is all good and worthy usability leading to better user experiences – but where’s the research contribution? Aren’t you just Doing The Right Thing? I know that’s rare, but that doesn’t make it research. Is there some theoretical contribution or generalization beyond the particular setting worked on?
• Very nice, but will it generalize? Please replicate your study with an n of 7 billion in all possible settings where computers are or might be. Then, maybe, I’ll believe you.
• Yes you’ve pointed something out. I agree that you’ve shown it happens. But so what? What can we do with that knowledge?
• Isn’t that just obvious?
• Won’t the problem go away if you just use {insert latest gizmo / business fad here}
• But what about all these other confounding factors that you ignore, or even state that you are ignoring? Don’t they swamp the effect?
• That looks like Smith & Jones (2009), a paper I’ve never read but has similar words in the title. It’s about computers too. Why didn’t you cite it?

Now on first glance we may think we’ve already addressed those issues in our paper. But sometimes they need to be made more explicit.

Why am I interested in this approach? Well because it is part of what I do when analyzing learning, confusions and misconceptions in my research. I think it’s possible to predict classic misconceptions about an application that some users can have that can be caused by or encouraged by an interface to that application. Then a bit of redesign can prevent or lower the odds of that confusion happening. Likewise I think it’s possible to predict classic misconceptions about a piece of research  that some reviewers can have that can be caused by or encouraged by a paper about that research. Then a bit of rewriting can prevent or lower the odds of that confusion happening. So just like user models and student models we can have reviewer models.

Think of peer review as a rather irritating kind of user test of your paper. It’s so much less fun than doing real user tests. It must be what it feels like when I test other peoples’ apps.

It can be hard to look at your own paper in this light.  Just like it can be hard to look at your own interface in this light. What I do is draw on the set of loony CHI reviewer personas I keep locked away in my head. If you try that, be careful that they don’t take you over, especially if you ever review anything I’ve written.

Presentation of Self in Online Life

Right, back again after a 14 month hiatus. No I wasn’t in prison, I was busy at work. But this year I’m on sabbatical and so it is a time to try out new technologies and think about how I and others learn them, how they get appropriated to new uses and how we might do redesign to Make Things Better. Currently I’m at UC Boulder visiting Profs Palen & Fischer:
http://www.cs.colorado.edu/~palen/Home/Welcome.html
http://epic.cs.colorado.edu/

http://l3d.cs.colorado.edu/~gerhard/

The EPIC group put on a local disaster just for me – very hospitable.

Now, I wonder, why did it take me so long to get back to blogging? Was it really that I was just too busy? Am I afraid of writing? Am I afraid of computers? They don’t seem very convincing reasons. Reflecting on one’s thinking is, I find, very informative – though you have to be careful about self delusion. Reflecting on one’s motivations is even more tricky.

I *think* part of this reluctance to post is more about that jumble of concerns about how one presents oneself in public. The Presentation of Self in Everyday Life  by Erving Goffman appears to be the book everyone invokes on this topic. I haven’t read it. I’ll add it to the ever growing sabbatical to-read list. But Wikipedia is as usual a great crib:

http://en.wikipedia.org/wiki/The_Presentation_of_Self_in_Everyday_Life

Also of course as a professor I have the luxury of outsourcing bits of my brain into a borg-like distributed cognition. I had the opportunity to work with Dr. Ben Gross on his thesis: Online Identifiers in Everyday Life. Ben read Goffman (so I don’t have to?). Anyway what I find really intriguing about his thesis was not the point that people worry about how they present themselves online in ways analogous to how they present themselves in Real Life. Rather his work shows how all those complexities can be bundled up in something as tiny as your email address. There are subtle degrees of prestige and meaning in what you choose and what you can get as an email address. Consider drunkenfratboy@stateUalumni.com, johnsmith6545324432432@gmail.com, ohmygodivehadthisemailaddresssiceiwaslike16-andnowitseemsabitinappropriateforacorporatelawyer@hellokitty.com

Ben notes “I repeatedly heard the phrase “I got my name” when people described why they signed up with a particular service”. I laughed when I first read that. Yes, smugly, I got my name in both Gmail and Illinois domains, but some random Twidale scooped me at Yahoo, dammit.

Having a rare or unique name is now something of an advantage in the internet world, and even makes up for all thos irritations of misspellings and mispronunciations.

http://www.cs.colorado.edu/~palen/Home/Name_Pronunciation_Guide.html

So if even something as tiny as your email address conveys a whole bunch about you, your importance, how early you were in getting in to the online world, how much you care about such things, just imagine the problems with Second Life and indeed WordPress.

Going into Second Life for the first time some years ago I had a body image crisis. Not just revisiting that endless teenage angst about “what should I wear? Will I fit in in a way that people will notice me – but not too much?” But also the virtual plastic surgery problem – shall I look like a 20 year old or try to look a bit more my age? Second Life does not do balding well. Bald yes. But bald as in “I could have a beautiful head of hair if I wanted to – this is a fashion statement”. Not bald as in “It mostly fell out except for a few odd clumps”. SL also does not do distinguished wrinkles well. You can aspire to look like Justin Bieber, but not George Clooney or Sean Connery. Hmm I wonder if there is a research project in avatar wrinkle rendering?  In SL white Tshirt and jeans subtly means “My Mom dresses me” (because those are the default clothes Mom-SL puts on new avatars). And we all know what that means.

So circuitously back to blogging. I’m now in that teenage girl at a new school / university mode endlessly trying on new wordpress themes. I can’t wear that blue thing you gave me. Just too dorky. Will the other bloggers take me seriously in brown? I don’t know. It’s all so complicated.

I want to go back to programming – no-one cares what a computer scientist wears in RL.

Thematic Confusion

I’m really confused by all this theme stuff. I like the way you can preview them. But most of them look just awful when I do that. Many shove a previous post of mine into the RHS column even though on the screenshots they look like they offer navigation bits. My current theme (Pressrow) at least doesn’t do that. But then it doesn’t put anything in the RH column. Rather the things I thought I’d added with the widget page to the RH column (Recent posts, Search, Email Subscription) appear at the bottom. Except on the about tab where they do appear on the right. What am I doing wrong?

Well I now see it works OK when you look at a single blog post, just not the big long list you get on the Home tab. Something very confusing is going on here. I suspect there is a blogging concept I’m missing.

Problem solved courtesy of this:

http://en.support.wordpress.com/themes/layout-issues/

Something to do with stray HTML. In my case I was using angle brackets in a non HTML way. It is a bit galling how much time I wasted before I came across this very helpful bit of support. There really does seem to be a big problem with how to search for help when of course you are confused and don’t know the right keywords. If only I’d googled: wordpress theme widget sidebar bottom. It seems so obvious in retrospect.

Tags and Categories

I wonder what tags and categories are and what the difference is.

Time to try out Help…

OK. That is pretty clear.

Post Tags are micro-categories.

Some more info: http://support.wordpress.com/posts/categories-vs-tags/

I got this from the wordpress feature of “possibly related posts”. All rather nice for incremental learnability.

Initialization screenshots

So I’m OK with uploading screenshots.

Let’s try a load of them. Here are some I made earlier…

Duh. Guess what, I clicked on the “Upload” text, not the funny allegedly-a-picture icon next to it.

Try again.

Lower half of that initial setup page, showing it is pretty small

So the default of full size seems OK.

When you click on “Next->” you get this:

The next page of the setup process

And clicking on “Signup ->” gives you this, if I remember right:

The activation screen

Then somehow you do the activation, via the emailed link, log in and get to the powerful but initially a bit daunting dashboard:

Oh. I must have clicked something wrong there. I uploaded it, but it seems to have gone into the gallery and not here into the blog posting. So how do I rectify that I wonder?

Lower half of that initial setup page, showing it is pretty small

The next page of the setup process

The activation screen

The initial dashboard

that first experience of a daunting dashboard

Trying to edit a blog posting

Problems with uploading images

after browsing for the image I just stop

well that wasn’t right. I think I inserted the whole gallery there. So what I think is going on here is that when you upload images, they also go into something called a gallery. But I don’t yet know what I did wrong. I must have clicked on something by accident when uploading  the last image so it didn’t end up in the blog post but in the gallery. And now I don’t know how to fix it.

Well maybe the simplest way is just to upload it all over again. Let’s try:

that first experience of a daunting dashboard

OK that worked. This time I managed to click on “Insert into post”:

Problems with uploading images

So I think I know what happened. I must have accidentally clicked on “Save all changes” instead of “Insert into post”. That takes you into that gallery thing which for me at my current learning state is all a bit confusing. No doubt it will become clearer in time.

By the way, as I’ve been busily uploading screenshots in this post I’ve noted I keep doing the same thing wrong. I click on the icon, browse for the image I want (the screenshot I’ve previously taken, cropped and saved), and then I just sit there:

after browsing for the image I just stop

That is, I just sit there waiting for the interface to do something, and it just sits there waiting for me to do something. We’ve reached an HCI impasse. It is waiting for me to click on Upload, and I am waiting for it to do the upload, because I’ve already told it what to upload. So why do I just sit there like an idiot? Indeed why have I sat there like an idiot something like 4 times in a row? I think it is that for me, the act of specifying the file I want uploading implicitly means “Here is the file I want you to upload”. And so clicking on the “Upload” button, which means: “And I want you to upload it” is rather odd. Obviously the system wants me to say “Yes, I really really really want you to do this incredibly dangerous and expensive activity and I appreciate we have to say it twice just to make sure”. And frankly I don’t want to bother. Just do it for heaven’s sake.

Furthermore, it’s not like the interface makes it easy to do one last check: “OK here’s this file you want me to upload. Is it the right one? Are you really really really really sure?”. No I’m not, because the text box by browse is so tiny, I can’t actually see which file I have specified, so I’m not given any clue that is is a ‘one last check’ kind of conversation. So I just sit there for a few seconds until I realize that this is the interface that makes me go “Duh”. A lot.

I don’t like interfaces that make me feel stupid.

Screenshots – second attempt

Right. New location, all firewalls off. Lets try Upload again…

the WordPress initial setup screen

So I think that worked. The previous problems I now blame on firewalls (somewhat superstitiously).

That one shows the setup screen that is pretty clear.

Upload was fine, though rather a lot of extra options I don’t want to think about for this first attempt. Still I think the defaults are OK, and they are pretty clear about what they do.

So there is that very nice simple startup screen. Though it isn’t actually a one-step form as it claims. It is simple and fast though.

So how do I add a screenshot?

Well it was all going so well. But now I want to add a screenshot to a post and I can’t see how. In the dashboard, the Media icon and link on the left seemed the likely suspect. That icon looks to me like a camera. But it seems to be all about sound and video. Nothing about static images like gif and jpeg. And it goes on about libraries which all sounds like advanced non bloggy stuff. So I wandered off and looked around elsewhere. But I can’t see any other more promising places.

Aha! now I see it. Maybe it’s in that row of icons just above where I am typing that says “Upload/Insert”. Yes, those words might be some sort of clue. I wonder why I didn’t see them before? I definitely saw all the text editing icons below them.

I see a musical note there, plus 4 other incomprehensible icons. Fortunately, hovering tells me what they are. The first is “Add an Image”.

So let’s click that and see what happens.

But first, that handy “Save Draft” button on the right had probably better be pressed.

Hmm. That upload didn’t work. But I suspect that’s a firewall problem of where I’m connecting online from. So I won’t blame WordPress for that. Their part of the upload non-process seemed pretty clear.

The mystery remains of why I failed to see that Upload/Insert text and row of icons. Maybe it is because the icons are greyed out. That would imply they are currently not usable. But it doesn’t persuade me why I failed to see them at all when I was looking for a place to upload a screenshot. After all I quickly found the (wrong) media link over to the left, and the icons there are also grey. And anyway the Upload/Insert text is not grey. Maybe I tuned it out as being above the (useful) text editing icons and below the (currently not interesting) permalink. But I don’t find that particularly convincing. It remains a usability mystery. In retrospect all I can do is rail along with so many designers: “How can he be so stupid/dense/unobservant? How can he not see Upload? It’s right there in front of him!” Such is the frustration of design – it would be so much easier if it weren’t for the users.

Creating that scary first post

More complex password duly entered and we’re off. A message telling me to expect an activation email in 30 minutes. So off to Eudora and hit “Check Mail” A mere 30 seconds later and it’s there:

To: twidale@illinois.edu From: “WordPress.com” <donotreply@wordpress.com> Subject: Activate twidale.wordpress.com

Howdy,

Thank you for signing up with WordPress.com.  You are one step away from blogging at twidale.wordpress.com. Please click this link to activate your blog:

http://wordpress.com/activate/7dd00a7675c9f1ac

–The WordPress.com Team

(If clicking the link in this message does not work, copy and paste it

into the address bar of your browser.)

Isn’t that nice! I’m familiar with this kind of email-click validation/activation. So I excitedly click and am told

“Your account is now active!”

So I log in and then hit my first surprise – this great big dashboard. Now if I knew more, I’d show you a screenshot of it here. But of course I’m currently a newbie and don’t know how to do that yet. I’ll add it to my to do list.

Anyway it is a big screen of stuff and info and stats. I’m used to this from web pages and other applications. But the really odd thing is that it appears that I have already made one post and got one comment. How odd! So I’m intrigued and click on it to see and it is a “Hello World!” post that in its content explains that it can be deleted or edited. Very nice. Though of course it is a bit geeky – if you studied computer science you know that the convention is to write your first program in any new programming language to say “hello world”. Worse, I had already just planned to title my first blog post “hello world” At last – mind-reading software. I’ve been waiting long enough…

So all that led to the previous post.

Good. Before too long I’ll be able to blog about interfaces other than blogging interfaces. I currently feel like I’m in the usability equivalent of the bootstrap problem.