Suddenly it's time for the third episode! Initially I struggled to think of new things to report on, but I can relate a couple of interesting events and instances of more gentle progress. The highlight of the past fourteen days was running the first statistics workshop. This was also the first time I've attempted to teach anyone about such a thing. I was very nervous but I had prepared good slides so it was relatively straightforward to get through. If you're interested, they’re available here. As you might have seen, this was not an advanced workshop. I started off with material from the first lecture of the First Year introductory stats course that I took, supplemented with material from an intro to stats webpage that I found. The workshop generated quite a bit of discussion at some points. When I gave the example of comparing eviction rates in the inner versus the outer khans Jim 1 and Jim 2 questioned whether a different outcome would have arisen had I selected different sets of khans for comparison. I didn’t realise it at the time, but this is an example of the modifiable areal unit problem (MAUP) which was made much of in my masters geographical information analysis textbook. Basically, if you change the size or shape of the area in which one is aggregating information for comparison, this may alter the outcome. This causes a problem for interpretation as the results may primarily reflect a particular areal unit rather than something about the aggregated data. I’ll raise that in the next workshop.

Another example of participants being engaged by the materials was when we learnt about the different scales of measurement. Typically when one thinks of measurement it’s about the temperature or altitude of something. Statistics, however, also involves measurements based on categories (nominal scale of measurement) and ranked categories (ordinal). I used a large dataset that they have collected over the past year, the 8 Khan Survey (of the 410 urban poor communities in Phnom Penh) to practice identifying the different scales. We found many examples of nominal and ratio data. The ratio scale of measurement accepts numerical data of equal units that does not go below zero (for example height). Interval data is the same, but can go below zero (for example altitude). There are a high number of fields in this dataset so most of the class was able to practice identifying a couple of them.

In the next workshop I hope to move beyond talking about concepts and get into the practicals from the subject reader that I brought over. Given that I am incapable of speaking Khmer and English is their second language, there is more chance they will get the concepts through practice. I think it was good, however, to start off with this introduction as the practicals will reference some of the ideas. Part of my work, then, in the next fortnight, is to decide precisely which practicals we will cover. There are 12 but at a rate of 1 a week this will take three months. It is unlikely we will need to go through them all to begin to come to grips with doing simpler things like correlations and chi-square tests. The statistical concepts that I mentioned having decided on in the last update may also need modifying. It’s a little difficult for me to know what to teach as for me it’s all fundamental material that should be covered before we even think about doing any analyses of our data. At least that’s based on my experience at university. It seems a little bold to chop out sections of a course but I suppose these are workshops after all.

Ok that’s enough about stats, although if anyone has experience teaching stats in a development context, please let me know! The next thing to report on is the five maps that have featured in the last two updates. I think I will stop saying they’re almost done, as something always comes up making them due for another revision. This time it is the dataset of villages excluded from systematic land registration. They have duplicate names and the only way to tell them apart is to categorise them by commune (Khmer: sangkat). This requires me to reformat the map as the table of villages has expanded with all the commune names. Now the table is too big to put on the same sheet as the map. This is good in a way as there is more page area for it, but it reduces readability as people will have to flick their eyes back and forth between the map page and the table page to find out what villages the labels on the map refer to. Another possibility is of course to increase the page size but this may cause problems for publication since the next size up is A2 and A3 is probably the maximum size one would want for a printed document that one is distributing. What I do know is we have 8 days to finish them off since we want to take them to the printers before the project coordinator goes on holiday. I think this will mean some overtime and a cut back on the cushy 1:30hr lunch breaks ... Especially since we haven’t even started on the separate maps of the four inner districts (Khmer: khan).

I also visited another three relocation sites. These were much the same as the first three although were slightly further out of Phnom Penh, in the range of 12-13km. One of them was very large, comprising about 1000 families. This one was very well planned out and the village chief was able to give us a photocopy of the site plan showing all the house lots and streets. This meant we didn’t have to survey it with our GPS. It had the air of an established suburb with solid looking houses and industrious inhabitants running shops similar to what one would expect anywhere else in Phnom Penh. The streets were all paved with tidy looking drainage down both sides. This contrasted with another, smaller community where there was a mixture of rendered brick two storey houses and wooden shacks. The village chief’s house was very neat although the rendering had not been painted. It was on a dirt footway rather than the paved road out the front of the chief’s house at the previous relocation site. This site did not have a bustling atmosphere and appeared to match more closely the stereotype in my mind of such places being less optimal for living than from where people formerly lived. I laughed on the way into the last relocation site. This had a ferris wheel and merry go round out the front! I thought it was a fitting challenge to the idea in my mind that these places were filled with oppressed, repressed people who certainly weren’t having any fun at relocation sites. Another interesting feature of the site was a massive quarry out the back. This was being serviced by a stream of dumptrucks that ambled along the eastern boundary of the site. As you can see in the photo, the quarry has turned into a lake. There was a rather pathetic looking pump barge/raft churning away at the bottom but I don’t think it would’ve done much. It was nice to get out of the office again and see for myself what the dots on the maps represent. There is so much information not communicated by a map. I think we should be emphasising our web map from this point forward as we can use multimedia communication to provide more detail on what the places are like.

This brings me to the progress seen in the Urban Voice web map. It’s come along really well. We’re thinking it’s the most advanced Ushahidi map yet deployed. The only shortcoming is the lack of SMS reporting facility, but that will be addressed in stage two. We have the first web map using the Ushahidi platform, AFAIK, that allows one to display more than one category/layer simultaneously. We also have the first one which provides embed code for people to place the map on their blog or site, like this. With these finishing touches there’s greater urgency for preparing datasets for upload. This will have to take a back seat whilst we push out the static maps mentioned above, but hopefully there’ll be a bit of progress in the coming fortnight. Another thing about the webmap that’s been in the back of my mind is the possibility of government censorship. A news blog called KI Media is blocked in Cambodia. It contains unbridled commentary on Cambodian current affairs. Every now and again I wonder if Urban Voice will gain similar attention from government censors, as we are encouraging people to express their views using the map. It will be interesting to see how things go, and I wonder if we might best judge the success of the service by the level of administrative ire it attracts … much like my respect for my NGO, Sahmakum Teang Tnaut, increased when I learnt they were suspended. It’s a funny way of doing business. And a funny way of doing mapping, given such professionals usually find themselves on the side of the government. Most who know me know I am incompatible with such a situation…

The last experience I will relate for this episode was very inspiring to participate in, but there was a twist. I went to a workshop run by Jim’s NGO Donor Organisation, a Northern Hemisphere funding body with money to donate to information communication technology (ICT) and mobile phone projects that encourage citizen participation in democracy. There were a number of NGO directors there as well as a representative from the ministry of education. The workshop revolved around a brainstorming session between participants about how the various NGOs might collaborate on a project worthy of Jim’s NGO Donor Organisation funding. The afternoon involved a session discussing more specific ideas stemming from the earlier brainstorming. It was wonderful to listen to what all of these very experienced participants had to say. I contributed a few basic ideas, for example emphasising that simple access to ICT/mobile technology was low in Cambodia and therefore we might want to focus on getting at least one computer with internet access for the public installed in each library. I also met four extra people engaged in GIS work of some sort. One of them had a GIS Google Group, to which he invited me and many of the other participants. This felt like great timing given that I am hosting a Phnom Penh GIS Meetup this Thursday at Gasolina.

So what was the twist? I wasn’t supposed to be there. You know how I mentioned there were lots of NGO directors? I am by no means such an individual... What I did was manage to crash a workshop created for what appears to be a specially selected group of NGOs that match the tastes of this particular funding body. I, of course, accept no responsibility for this whatsoever. I was merely there at the behest of my boss who had made a quick decision the preceding day to have me go along since she didn’t have the time and it looked interesting when the director of our web design contractor mentioned he was going to it. Assumptions were obviously made that shouldn’t have been about whether it was invite only, and I didn’t think to question them as I was simply told I was going and who I should engage with there to spread propaganda about our web map. I was successful at this, talking with almost every participant about the site and distributing our bumper stickers… but how I cringe now knowing I was oblivious to the fact that I was uninvited! So how did I come to know? My enthusiastic email asking whether we might be able to meet with Jim’s NGO Donor Organisation next week was met with the straightforward reply that they were delighted at my response to the workshop but were perplexed as to why I came as a representative of my NGO was not expected. Why they didn’t just tell me at the start that I couldn’t participate is a mystery. Perhaps I came across as undeniably earnest. Despite this possibility, they refused our request to meet them this week. There are, of course, no lessons to be learnt from this experience.

Making trouble in Phnom Penh

My motorcycle is going well, although registering it is proving to be a pain. I have two avenues. I can register it through my NGO and get a special NGO plate, apparently. This reportedly causes the police to give me less attention. This would be good, but I wonder if my NGO is too small for such grand privileges as special number plates for staff. The other option is to have a local register the bike in their name, then transfer registration to me using a standard vehicle purchasing form. I think it’ll probably end up like this. I have ridden the bike little during this period as, being the conscientious Westerner that I am, it feels terrifying to ride an unregistered vehicle on public roads. Last weekend was a lot of fun, I met a lovely bunch of English teachers, one of whom is especially, particularly lovely. This one is from the US, which is providing some good opportunities to finally get some details on life in the world’s latest empire. Swing class was horrid this week, we did the Charlestone which just made me feel like I was always going to fall over. At least they weren't try to get us doing this. I’ll need to do more practice. I have also moved house. Now my room has a door, bed, air conditioning and an ensuite! The rent’s tripled but it’s still a share house and is close to work and friends. On Saturday there was a full lunar eclipse. Audaciously, my phone camera decided to try to use the flash when taking a picture of it. The easiest way to describe it is that it was like a normal full moon but less bright and red tinged. Last of all, as promised, I finished the Angkor Wat Bike Race, fundraising for Village Focus International. It was 100km of flat, basically square, potholed asphalt which I did in 3:40hrs on a mountain bike. Thankfully with the forks locked out. I found a good bunch to ride along with and all seemed to have good riding experience as they were happy swapping the lead and drafting. It was very nice to relax in Siem Reap afterward just letting the day go by.

If you’re in Phnom Penh and have an interest in maps, get along to the Inaugural Phnom Penh GIS Meetup! 7:30pm onward at Gasolina, Thursday. Check the flyer below.