I think we can put this point on our skype/IRL agenda
I see the dilemma, but did anyone seriously try speech2text yet? Even if it has hiccups (that will likely improve over time) it may be good enough for the purpose.
Now weâre talking!
I think this is the right direction. I would suggest a few things to consider:
90% of video content runs in the background for me, and I hardly watch it cause it rarely carries much information and warrant precious âeye-timeâ. Thus I would suggest splitting the two things:
You hand out a bunch of recording devices (ideally something like the olympus ws-801, realistically cell phones you dangle around your neck and record in amr). Meanwhile one of you jumps around and makes short videos of revelant stuff + tiny intros of the people to have a face and match the voice to name/email (perhaps written on cardboard).
Integrating with the ER platform can also happen later, but would be nice. From my perspective getting the raw-on-the-ground-info+emotion accessible is the main thing. I would also let them record in Nepali if that is significantly easier for them.
You could also use the soundcloud account I made (is good for 180 minutes). Iâll send you the access info. It should be possible to link to a specific time from the ER platform.
2 Likes
TranscriptionâŚ
@trythis, have you seen the exchange between Matt and me above? I know you are a fan of audio, but how would you deal with the issue that we are committed to handing out an ethnography?
Long storyâŚ
Iâve seen your exchange and you can probably imagine that I am cringing pretty bad here. I see your desire to deliver a âcleanâ, relatively conventional ethnography - and I empathize with it. But I think it would be better to widen the scope in a way that hopefully will not compromise but rather complement and support the text based facet of the ethnography. I would feel much better detailing this in a slightly intoxicated (the Greeks new what they were doing) heated 3-4h discussion, with my mirror neurons firing. Iâll also offer to do minutes for it, and perhaps try to speak without too much of a slur for an automatic transliteration attempt.
Creative ways for written online dialogue
Some initial ideas how to organize the online dialogue here so that participants donât see it as âexcessively challengingâ but rather as âworth the effortâ when contributing in writing:
- SMS stories challenge. "Tell us what you do in disaster response in 10-20 SMS and get invited to our workshop in two weeks." We have a lot (100-200) mobile phone numbers of people engaged in the local disaster response initiatives, and could send them a SMS invitation.
- (to be continued)
Double pack
Perhaps you could do an honor system where they are asked to record one now, in return for also transcribing one of someone else. You could say they should perhaps rehearse their story once to make sure it is below X minutes, and there arenât many extended silent stretches. Maybe do a 30 second summary. That could also help improve the quality (slow thinking) of the contributions - but would move away from a clean response.
You could offer help/the infrastructure for the transcription and at the same time get a chance for a follow up meeting.
Dialogue=Conversation
Hi there, I think we need to consider few things here:
-
The relief teams/initiatives are usually busy managing and coordinating relief work (a friend of mine has been on the run since the earthquake) which means the time that they do the reflection is probably on buses, micros, or during phone calls sharing information/story/updates. And as much as I would love for my friend to write on the platform, and start engaging in the space, he simply does not have the time to.
-
But, there are relief teams that are coming to a closure of some sort - perhaps, identifying these groups and meeting them first-hand to explain what we are trying to do would work better. I have messaged, talked to few about this and yet I think it would be a more compelling case to have a face-to-face conversation and then let them decide. The first barrier to online dialogue is it takes a while for people to get used to a new interface (if it was Facebook, I am sure there would be more users!). [Also like Matthias pointed out, we have so many of such conversations every day we meet someone who is a part of the relief team.)
-
About videos/audios/transcripts/SMSs: I love the SMS idea. We can definitely give it a try. I think as far as I understand the discussion, we want first hand narrative and I completely understand why. I think we will need to work on some tasks and start spreading the word. I havent yet shared much about by involvement in this project in my own Facebook/Twitter accounts - and once we start generating some kind of activity there, it will hopefully channelize some of that into this platform.
-
I slept over these comments - and I think the only way we can have dialogue online is if we are in contact update/loop about these projects ourselves. If we keep in touch with the people through Facebook, emails, and even meet, I think maybe they would be interested to post in this platform. I already see quite long updates/descriptions being posted on Facebook about such and such relief work they did. We just have to redirect some of those into this platform. (I also say this because relief teams are working on âbatchesâ so they take a team, go, do their deed, and come back for a day or two - and then repeat that cycle. So maybe they will be free on those days - rest, and also reflect?
-
I prefer audios for first interviews because that puts less pressure on the speaker to âperformâ (some are natural though). We can encourage them to share videos they have taken, photos they have taken though. Maybe in one of the workshops, we can also learn/share about visual documentation and they can make their own?
2 Likes
Challenges versus Opportunities
Hey guys,
I was hooked into the conversation before I knew it. There are suggestions flying and they all look good. I have a few points:
-
I think, the best way to get people to talk is to talk to them in person. That I tell you from my experience as a journalist. It is much easier than online discussion in Nepal. It was true before the disaster and today, it is more true than ever.
-
People are more comfortable in GROUP DISCUSSIONS. I recommend a group of 5-6 people (preferably friends because it will be easier to manage time) with us prepared with questions revolving around the theme. The challenge here is not lack of resources (people). It is getting the time to make them talk.
-
I like the SMS system. I also want to add that there is a time constraint. How about we prepare the questions we want to ask people. If we want them to âtalkâ then they do. But weâd probably be able to take 20 percent of what they tell us. People sometimes prefer a guideline. This goes into email exchanges as well. If we define objectives and come up with questions, then we can forward the emails to people and probably get more responses.
-
I agree with Dipti when she says people are conscious in front of camera in Nepal. Audio is a much accessible means. Both Dipti and I are experienced as journalists so we can even note down the questions (supported by audio of course). We can have few people talk in the video but most would prefer audio (probably with a picture of the speaker?)
-
Finally, I donât know how much time we have (and I hope Dipti agrees to it) but we might get the richest resource when we have in-depth discussions with the people. I want to do a few blog posts where I talk to few inspirational people working in the field.
-
We can use social media more actively. How about we use the same introduction we have ready in all our personal Facebook Profiles and then invite people. I donât know how much that can help but we can try.
Again, any suggestions are welcome. Dipti and I are meeting today to discuss it further. We will keep you updated. Meanwhile, we are trying to get in more people the old way as well! 
3 Likes
Crowdsourcing answers to local activists questions?
So hereâs my thinking. With time and connectivity problems, people on the ground will be having a hard time accessing support from outside the country. At the same time, people who care outside might be looking for ways to support their peers in the Country but at a loss for how (especially if they donât know who/which actors are trustworthy).
Do you think it is feasible for the futuremakers team to convene, facilitate and document physical meetings where people are discussing their hands-on experiences trying to do different things and trying to make sense of how to deal with different challenges? If you get peopleâs email addresses and their permission to submit documentation of their own words on the platform for them then perhaps members of the Edgeryders team in other parts with less problematic connectivity could manually input the content onto the platform.
If so we could fire up the agency and crowdsource input from the broader community and synthesise it into suggestions. Possibly in the form of a crowdfunding campaign where people around the world can contribute knowledge, time and or money and the proceeds go to supporting people and groups on the ground. I had been thinking about this approach for the Future Makers project teams in Armenia, Belarus, Egypt, Georgia, Morocco and the Ukraine here. So we could try and push this as a global campaign to support futuremakers in different parts of the world at the same time and the ethnographic input goes into supporting this process.
What do you think?
2 Likes
I like this, but Iâll try to be contrary
From a âscientificâ point of view:
Iâd prefer the sources to talk freely, without much interaction with the interviewer. Even if you donât say a word, your face will communicate â10x the amountâ of what e.g. goes through this text box. So thereâs a chance we influence the results and they would become less representative. Writing a text âalone in a dark roomâ is far superior in that respect. So Iâd suggest you practice a little on how to do the intro, and perhaps leave a questions written on a folded up piece of paper with them. Then you first let them do a few minutes of free response, then they open the first question, read it, think about it for say 3-5 minutes (you want them to think, you donât want them to babble), and then do a 3 minute reply for each, and close by giving 2 contact handles. You could jump from one source to the other with a few recording devices or (if you think it is better) attend the recordings if you are not too busy. I think this is best done with an interviewer and one assistant to help stick with the protocol.
I would be very interested in what groups have to say as well, BUT: In groups there is LOTS of social dynamics going on, which makes this a totally different ball game from 1on1s (or 1on0s). This is not necessarily bad, as we are group animals and most of what is done is done in a highly social context. My (perhaps wrong, and surely anecdotal) experience is that the people least likely to pipe up are the people with the most insightful comments (introverts are by and large more sensitive, which allows them to factor in weak signal better). [edit: she would probably be interested in the results as well, + I think it is also a gender thing] Yet I would refrain from trying to moderate them into talking as it would make the reaction very unrepresentative (in my view). What I would do is make sure you get 1on1s AND groups so you can also show that they are different things. Ideally youâd get some (random, probably not possible) of the same people into 1on1s AND group discussions.
I would also recommend (at least in some randomly determined interviews) to pass the mic around during group interviews and not have everyone babble at the same time - cause that would make transcription total hell.
I think you are right with the guidelines, but itâll probably takes us 1-2 iterations before we get a good one. And we want to be consistent from a scientific method point of view, I think. So before that happens I would say we should purposefully pursue a less structured and more flexible approach (always with the goal of making such a guideline in mind). First the data, then the categorization - in complex issues (almost anything involving humans is complex to a degree) Ideally one would try to capture the meta-level as well, i.e. how do you make a meaningful guideline. There must be resources on that online, but I am not sure if theyâll work in our context and what the search terms would be.
The sms is probably a really good idea from a âcleanliness of dataâ perspective. How can we get people doing that more - because I am afraid itâll be not utilized a lot. Flyers? Incentives (may compromise data quality)?
I would offer picture or tiny vid (perhaps with a little cardboard sign with a witty phrase) as an option AFTER the rest is in already. We donât want to lose our introverts. 
Perhaps there is a way to suck in the people that have responded in the âlow threshold modeâ at a later point in time? If you already invested time in something, you are more inclined to do so again - even if that is less fun. Here, I would do a two pronged approach: Do it âby the bookâ, selecting really random people from the list. And go âmindfully cherry-pickingâ - this means you look for interesting outliers (weak signals) and milk them hard for more data - that could perhaps support an unorthodox hypothesis. BUT BE MINDFUL! This means being aware that you are your greatest enemy most of the time. And certainly when you donât think so.
Individual outliers (as you rarely get them in group interaction) could then be invited to a more extended group interview and guest posts.
I think we should do a question dump where people from far away (like me) can ask their stupid and in-sensitive questions. Also it would perhaps interesting to have a sort of âmessy rankingâ what the people on the street/field in Nepal have on their mind, and how urgent they perceive which question.
Lastly in all my enthusiasm for audio (ahem, I HAD said I wanted to be contrary in this post), Iâd want to urge people really hard not to neglect the text based format and inadvertently âdry it upâ by opening âchannels for lazy peopleâ that end up sucking everyone in. I think this would actively do harm to an important part of work that @Alberto and many other have devoted a lot of effort to. So we need to come up with ideas that strengthen text, and allow people to migrate from audio into text, at every corner. Yep, I know I havenât been leading that charge very effectively so far. Sorry.
Scientific, of course
Yes, @trythis. The added value of ethnography is that it incorporates the point of view of the group being studied. Speaking in the first person is important, and we should take great care not to lose that advantage. However, the Nepal team has its own resident ethnographer, @meenabhatta; I expect she will be the guardian of scientific propriety and will stop us from making too much of a mess. What do you think, Meena?
(for the same reason ethnography was always a small sample discipline: you could do legitimate studies with 15 respondents. That we tend to have 100-200 is made possible by⌠online ethnography)