Hi Yiorgos,

thanks for your email! Please find few comments in line


> Why is there not a possibility to enter into the survey a little bit more data regarding each participant's 'urban visibility reach' of the city (the space of the city that one is actually able to better know visually)?

if the number of participants is high, then we would need to control
as we are dealing with a randomized study. if it isn't very high
(which might well be the case), we need to control for it - that is
why we asked for postcode. since we need to keep the final survey as
short as possible, we could modify "postcode" with "the area(s) you
know the most (comma separated postcodes)". so that the survey length
would be the same and we would capture what you've suggested. what do
you reckon?


> It appears to me that the 'recognisability' of a city by a particular individual would depend among other things on the ratio

you are right. the question is though: is recognisability (which
depends from a zillion factors) a good *proxy* for well-being? to
answer that question, one is after correlations not causal
relationships.


> "the first thing I read in [an image] was the typology of tube stations of a particular line"

i'm not sure whether we could somehow account for that ...


> Are the results of the quiz treated according to the possible specific clues that each picture contains?

Yes. Secret: some pictures have been chosen to be very recognizable
and are not counted in the study (they are fake). they just serve to
increase retention rate


>>> Daniele,
>>>
>>> very interesting. One aspect that I suspect would be useful in thinking this through would be to have some putative mechanisms to hand. In particular Kate Pickett's observation seems to me to be telling. I wonder why she wouldn't recommend looking at inequalities at the local level? I suspect that this is because at that level it looks to her like a mess, whilst at the level of whole cities or regions she can see some patterns. This seems to me to be a matter of the lens one looks through, and perhaps she is not looking at things locally enough.
>>>
>>> I regularly use an image from one of Charles Booth's maps of 19th century London to illustrate the point that social class as he mapped it seems to be closely associated with urban space at the street level. As you travel along a street the colours (classes) tend to be the same on either side of the road or only to change slowly, but as you turn a corner they change, turn again and they can change again. The result is that within a local neighbourhood (take an administrative boundary such as a Ward that Kate might consider to be 'local') and yes the whole thing looks like a mess - all the colours are there and when averaged they may be pretty meaningless. But consider it even more locally at the level of the street that we walk along and experience, and there is a strong pattern. This pattern is correlated to other aspects of what we experience - scale of the buildings, level of investment in the architecture of the facades (your survey has some nice examples of this), traffic flows and noise levels, the presence of pedestrians, property uses, and so on.
>>>
>>> The 'space syntax' counterpart to Lynch's notion of 'imageability' is that of 'intelligibility' defined as a correlation between local properties of space and global properties. How well, from what you see locally, can you predict where you are at the large scale. While we calculate this purely on the basis of correlations between network measures (e.g.. Connectivity:Mean Depth). However, local properties for me include many of those described above, but traffic flows, pedestrian movement and as a consequence of these land use distributions (in the long evolved historic city and the un-planned settlement) depend upon global measures of network accessibility in urban space. We can see them locally, but they depend upon the global structure of city space. As we walk the streets and 'learn' the spatial culture of a new city I suspect that what we are learning is the association between local and global. This association we then use to help us navigate and way find. It is something that we learn anew each time we visit a new city or culture. It is possibly one reason we enjoy being a tourist.
>>>
>>> Sheep Dalton did a very nice set of experiments in his PhD that are relevant. There is a body of literature that suggests people's attachment to 'place' is an important part of their identity. Sheep asked people to identify on maps the boundary of their local area. He then developed a measure of intelligibility that could be quantified for each line in an axial map of a city. Interestingly, this measure shows a patchwork across neighbourhoods that has a surprising degree of agreement with both named areas and with the boundaries between localities identified by people in the questionnaire.
>>>
>>> I suspect that the relationship between intelligibility, place and individual attachment/identity may be a part of what you are trying to get at. It may be the lack of intelligibility correlations resulting from functional zoning and more recent planning processes that creates the kind of anomie which might detract from well-being. Perhaps underlying this is a kind of a mechanism to be tested against the statistics?
>>>
>>> Sorry about the length of this.
>>>
>>> All the best,
>>>
>>> Alan
>>> On 24 Feb 2012, at 09:44, Daniele Quercia wrote:
>>>
>>>> Hi Alan,
>>>>
>>>> Thanks for your reply. I agree with you. The well-being bit tends to
>>>> be quite complex, yet there might be a certain sense of direction in
>>>> the literature, to different extents depending on the unit of study.
>>>>
>>>> a) If the unit is individual, then research has been extensive. There
>>>> are widely recognised tests to measure individual well-being (the most
>>>> popular of which is the satisfaction with life - SWL - test), which
>>>> goes well beyond asking people how they feel (which is generally only
>>>> done in experience sampling studies). Daniel Kahneman's latest book
>>>> [7] brilliantly summarizes a variety of effects (e.g., memory,
>>>> adaptation effects) associated with  well-being studies. I don't know
>>>> of any meta analysis in this area, but based on what I've read (which
>>>> includes [1],[2], [3], and [4]), I feel that there are peer-reviewed
>>>> and validated tests to measure individual well-being and, more
>>>> importantly, results about individuals of different nations and
>>>> cultures are consistent (e.g., the u-shaped function with age,
>>>> importance of social relations, tail-off function with income).
>>>>
>>>> b) If the unit is community (say UK census area), then one either
>>>> administers SWL surveys to a representative sample of each community
>>>> or  uses proxies for well-being (Nathan and Michael have used IMD at
>>>> UK census area level [5], some UCL folks are trying to come up with
>>>> different composite measures based on a geographical partitioning
>>>> different than that created by census areas). However, even if one
>>>> were to administer existing surveys to everyone, then the problem
>>>> would be that these surveys do reflect individual well-being but not
>>>> *community* well-being (which might well be different than the
>>>> aggregate well-being of the community's residents).  The only effort
>>>> in this direction seems to be that undertaken few meters away from
>>>> where Kevin Lynch was teaching - Somerville, next to Cambridge (MA)
>>>> [6].
>>>>
>>>> c) If the unit of research is nation, then metrics that have been used
>>>> include Gallup-Healthways Well-Being Index (at US state level) and
>>>> income inequality (which, at nation level,  correlates with life
>>>> expectancy, mental illness, etc. [8]). However, taking these metrics
>>>> at a different spatial resolution (say UK cities) would be plainly
>>>> wrong. One year ago, I had a chat with Kate Pickett on this and she
>>>> said "We would not recommend measuring income inequality at a local
>>>> area level, see paper [8].  Social capital is, of course,
>>>> appropriately measured at such a level.  For information on UK data,
>>>> see this website:
>>>> http://www.ons.gov.uk/about-statistics/user-guidance/sc-guide/index.html".
>>>> I also feel that Richard Layard of LSE would have something to say
>>>> about complementing (not replacing) traditional economic indicators
>>>> with well-being indicators.
>>>>
>>>> As for interaction effects among different factors, they are easier to
>>>> control at individual and nation levels based on cross-lag analyses
>>>> (because of decades of research  in economics), but they are less so
>>>> at community level (because of few isolated & recent studies).
>>>>
>>>> Sorry for my long email :) Thanks again for your input.
>>>>
>>>>
>>>> [1] Richard Layard. Happiness. Lessons from a New Science.
>>>> [2] Ross Gittins. The Happy Economist.
>>>> [3] Jonathan Haidt. The Happiness Hypothesis
>>>> [4] Nick Powdthavee. The Happiness Equation
>>>> [5] Eagle and Macy. Network Diversity and Economic Development.
>>>> [6] Somerville, MA. A report on Well-being
>>>> http://www.somervillema.gov/departments/somerstat/report-on-well--being
>>>> [7] Kahneman. Thinking, Fast and Slow
>>>> [8] Wikinson and Pickett. The Spirit Level.
>>>> [9] Wilkinson and Pickett. Income inequality and population health: A
>>>> review and explanation of the evidence
>>>>
>>>>
>>>> - daniele
>>>> http://www.cl.cam.ac.uk/~dq209/
>>>>
>>>>
>>>>
>>>> On Thu, Feb 23, 2012 at 6:42 PM, Penn, Alan <[log in to unmask]> wrote:
>>>>> Daniele,
>>>>>
>>>>> I think that research has gone quite a way forward on the measures of visibility, imageability and intelligibility side of things (not least in the space syntax field). GIS has done wonders for being able to get hold of data to analyse and there have been many different relevant approaches, work on viewsheds in landscapes etc. including a number of people looking to reproduce Lynch's findings. A lot also from cognitive science that I would think is relevant.
>>>>>
>>>>> I am less familiar with the 'well-being' side of the equation, but suspect that this must be much more difficult to pin down - it must be very multifactorial (even more so than 'health' which is notoriously hard to handle). My guess is that the whole thing will be hedged around with all sorts of autocorrelations - wellbeing, relating to health and to inequalities, both of these related to poverty, wealth and education - all in interesting and complex ways some going in counter intuitive opposite directions. e.g.. people dying of cancer saying that they feel 'well' versus people with head colds saying that they 'at deaths door', poor people being happy and rich unhappy etc. All also strongly patterned spatially so some real underlying issues of cause and effect in trying to unpack any association. There is a lot of work going on on this of course - Cameron's happiness agenda, Marmot's health inequalities etc. but I have not heard of anything approaching a consensus or even a sense of direction.
>>>>>
>>>>> All the best,
>>>>>
>>>>> Alan
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> On 23 Feb 2012, at 16:37, Daniele Quercia wrote:
>>>>>
>>>>>> Kerstin, thanks for sending the email out. And thank you all for
>>>>>> playing the game!
>>>>>>
>>>>>> We are modifying the site as I (literally) type ;) We have added a
>>>>>> description of the project and will shortly add commenting features
>>>>>> (thanks, Irene)
>>>>>>
>>>>>> We have also downplayed the link between well-being and
>>>>>> recognizability. Thanks, Alan. It would be great to know whether
>>>>>> research has gone beyond Lynch's book "The image of the city"
>>>>>>
>>>>>> A full description is here
>>>>>> http://urbanopticon.org/release.html
>>>>>>
>>>>>> Any other comment is welcome!
>>>>>>
>>>>>> - daniele
>>>>>> http://www.cl.cam.ac.uk/~dq209/
>>>>>>
>>>>>
>>>>>
>>>
>>
>> --
>> The Open University is incorporated by Royal Charter (RC 000391), an exempt charity in England & Wales and a charity registered in Scotland (SC 038302).