I've been feeling guilty lately that James has been doing all the heavy lifting on the blog, particularly writing about the Kidderminster meeting and the analysis, so I should probably say a little bit about what I've been up to.
At the last few presentations we've done, James has talked about the analysis essentially being on two levels - the first being quite schematic, number crunching stuff and the second being the clever textual analysis of meaning etc. So while he's been working his way through the transcripts coding things, I've been painstakingly fiddling with the GIS.
The first thing I've done is to think about the places where people have walked and try to break this down into categories. Clearly, when you walk in a city, you can't just wander where you want and you tend to follow linear features of one kind or another - particularly roads. So I produced a list of the different types of features that appear in Eastside:
1) Primary distributor roads (e.g. Digbeth High Street)
2) Secondary distributor roads (e.g. Fazeley Street)
3) Tertiary distributor roads (i.e. any of the back streets)
4) Canal tow path
5) Paths (i.e. pedestrianised and 'other' areas)
I then looked at the areas where our interviewees have walked and redrew the map of the area, categorising all linear features into these five different types (I've posted a working model of this in the Downloads section of the RG website).
One of the problems of the GPS tracks is that they tend to wander a little bit - if you look up close at any of the tracks on the website, you'll see they often pass through buildings etc. because the accuracy is at best around 6m and, particularly around the viaducts, often quite a bit worse than this. So I've used the tracks to draw rough 'corrected' tracks, which could then be split into pieces depending on what kind of linear feature they were passing through. This allowed me to calculate approximate distances walked within each feature type and, again working back through the GPS logs, how long was spent on each piece of the walk.
What's the point of all this? Well, one of the things we were interested to find out in analysing the walked interview method was whether external environmental factors had an impact on the way people walked. Do people, for example, avoid noisy roads and, where they have to pass along them, do they walk more quickly? Well, with all of this data about different road types, times/distances we can work this out. And the answer is... well, sort of. There does seem to be a bit of a tendency to spend longer walking the same distance on tertiary roads than primary. But the time/distance ratio is pretty similar for primary roads and secondary and people seem to pass most quickly at all along the canal towpath. This might be because there's less to stop and look at, but we're going to have to look at James' content analysis data to unpick that a bit more clearly.
The other thing I've been doing is taking the hourly recordings from the University of Birmingham's weather station at the Botanic Gardens in Winterborne - just over a mile and a half away from Eastside. I've plotted the dry bulb temperature, windspeed and level of precipitation against the length of interview, to see if there's any discernable pattern. There's nothing especially obvious, though it's reassuring to note that the one occasion where I definitely know it was hammering with rain (Jane's interview with Julia from MADE) is the one occasion where we have rainfall recorded at Winterborne - 0.2mm in the course of an hour, which a meteorology colleague assures me is quite a lot.
So, essentially, a lot of what I've been doing has shown somewhat ambiguous/negative results, but it's good to be generating this kind of data.
This said, however, I do have some concerns that we've ended up doing some rather positivistic science i.e. reducing people to a bunch of statistics. Particularly as those statistics are coming from the rather big brotherish surveillance technologies of GPS and GIS. I think the tension between this kind of approach (and the inherent power imbalances it contains) as against the rather more fluffy aims about empowerment that we have in mind for the 'community' side of rescue geography are things that we're going to have to unpick at some point.