Experimental confirmation of broken windows "theory"

[Adrian Short]

Hi everyone,

The Economist has an interesting article on research at the University
of Groningen that attempted to confirm or refute the hypothesis that
environmental conditions of "disorder" affect people's behaviour:

http://www.economist.com/science/displaystory.cfm?story_id=12630201&CFID=31056247&CFTOKEN=41038121

I've done something similar unscientifically and informally through
painting out graffiti in an alleyway in my neighbourhood. It appeared
that the less graffiti there was, the slower more graffiti would
appear.

My own thoughts on the issue are here:

http://suttonlivingstreets.org.uk/2006/04/16/8/

--
Adrian Short
http://adrianshort.co.uk/

[Dan Lockton]

Thanks Adrian,

That's a very interesting study. I know there are plenty of criticisms
of the broken windows theory as an overall predictor of other kinds of
(more serious) crime and I reckon you hit the nail on the head with "A
neighbourhood that clearly says, “You can get away with graffiti here”
doesn’t translate as, “You can get away with murder here.”" in your
blog post.

The thing that interests me particularly about the 'single broken
window' aspect of the theory is how it seems almost like there's a
discrete jump from 'totally clean/pristine' to the category of damaged/
broken/grafftiti'd/littered etc, which we pick up on first and use as
a cue which then influences our perception of what actions are
appropriate (what the norms are).

That does appear to hold true in other areas of life, at least for
some people. A door is either closed or open, and once it's open it's
easy to open it further. Once your desk starts to get cluttered, if
you don't tidy it up, the rate of cluttering accelerates. If you don't
oil that squeaking door hinge when you first notice it, you won't
bother oiling the next one either, since it doesn't seem like the
problem is any worse, somehow. It's almost like Boolean algebra:
"messy + messy = messy", once messy is above a kind of threshold. This
seems to be part of the idea behind one aspect of the 'Getting Things
Done' productivity method - http://en.wikipedia.org/wiki/Getting_Things_Done#Principles
- if you can deal with problems or at least work out properly how you
are going to deal with them as and when they occur, you'll save a lot
of work and stress later. Easier said than done, of course.

What are the implications for design in general, then? Does a clean,
uncluttered interface for a product, service, or a building, or indeed
a person, translate into a kind of micro-perception that 'care' has
gone into it (and thus it is appreciated more?) Conversely, does a
messy, cluttered interface imply a lack of care (and thus lower users'
estimations of its worth)? Or is that pushing the theory too far?

Is this whole field actually a kind of design/architecture/behaviour
variant of the fundamental attribution error (http://changingminds.org/
explanations/theories/fundamental_attribution_error.htm ) in that
we're extrapolating general characteristics from some specific
details, and thus changing our behaviour accordingly even if it's
largely irrational? E.g. we see the graffiti and assume that the
general behavioural norms in the neighbourhood are different, hence we
adjust our behaviour towards them?

If there are two identical second-hand cars for sale, priced
identically, but a pigeon has done its business all over the
windscreen (and it hasn't been cleaned off) of one of them, will
people be willing to pay more for the clean one, just because it's
'more cared for'? Will they assume that the dirty one has other
defects?

Apologies for a long and rambling post, but it's a great subject -
thanks for the link.

_________________________________________________________________
Dan Lockton MPhil BSc(Hons) FRSA | Cleaner Electronics Research Group
Brunel Design | Brunel University | London | UB8 3PH | http://danlockton.co.uk