Monday, August 19, 2013

Mobile ethnography tasks and probes:






















The single most important part of any mobile ethnography project are its tasks. By which I mean the questions and more often activities you need participants to video, photograph, audio record or note down.

Below follow fifteen tips which will help with your next mobile ethnography project.

Practical advice first:

1) Whatever your tasks, always ask you mum, sister, husband, cousin (not someone close to the project) to read them. Do they understand the tasks? Are the tasks open to interpretation? Do you want them to be? Especially important with multi country projects and languages. And never assume your tasks are so simple that cross checking is not necessary.

2) Is the participant an observer/camera person filming other subjects or will the only capture their own lives? Sometimes it makes for better ethnographic research to have someone else filming you cooking, completing a DIY project or applying makeup.

3) There are a thousand great reasons to create sample responses to tasks for participants to view/read in advance of capturing their own tasks. A sample video, photo, audio or text task, filmed/photographed by you, will create a quality and consistency benchmark for participants. Create one and let participants understand or 'get' how you want them to hold the camera, take the shot, describe the event.

A sample task also allows, you, the project manager, to experience first hand what it will be like to capture the tasks. Enabling you to make improvements to wording or even change the task itself.

Worried you might influence what they capture and language they use by seeing your sample responses? Don't be. You can also use other participant entries - more than one - as a quality benchmarks by sharing with other participants as your project progresses.

4) Do you have a preference about what media your participants use to capture an event? Be clear about which media they may use. Don't assume the type of the task will make it clear.

5) You may have read this in my previous posts but still worth repeating. Don't ignore the power of audio entries. People love talking. The difficulty with videoing and talking/narrating at the same time is that you stop when you run out of things to film. A straight audio task, perhaps using an existing entry/event as stimulus, can make for a very rich entry.

Designing and interpreting tasks:

1) Break fieldwork duration into two parts: a) capture stimulus and b) reflect. Start with simple, quick to complete activities and end with more reflective probes using the earlier, quick entries, as stimulus.

2) Don't turn the project objectives into questions/probes. This may sound obvious, but I regularly see awful sentences such as: "How does xxxx fit into your life/day?". Participants shouldn't be asked to do your job. You need to interpret photos/quick videos of each time they consume xxxx and draw your conclusions once you have used the same photos/quick videos as stimulus for more reflective tasks/probes.

3) Tasks should be designed to work 'in-the-moment'. Don't ask people to explain choices when in the middle of a supermarket shop. Nothing can be more contrived and disruptive. Instead ask them to snap/video away so you can probe/explore later, with a new task.

4) If you want to ensure complete naturalism, disguise the brand/product/activity/service you are interested in by asking them to capture a range of different activities or brands. You can reveal the brand/product/activity/service you are interested in towards the end and then probe away.

5) Tasks need to be varied, interesting and fun. Boring tasks = boring entries/outputs. A good way to think about a mobile ethno study is as if you are inviting a bunch of friends to a party. Make it fun and let them have fun, on their terms. In other words, don't ask them to set anything up such as a fake shopping trips.

6) Take the plunge and use yourself as a benchmark: "I like xxx for breakfast what about you?". Participants who connect with you generally (not always) provide richer content.

7) At then end of a project, always ask participants to go through all of their entries and tell you which and what surprised them most about their captured behaviours.

8) As above but with another or group of respondents. Different perspectives will help to add new meanings to same old events.

9) A key deliverable in ethnography is 'things you didn't know you didn't know'. And I was quoting this at conferences way before Donald Rumsfeld ever did. An effective way to achieve unknown, unknowns is to invite friends or family members (of participants) to review entries sent by a participant. We often get comments sent from family members to participants such as, 'Since when have you been drinking water/using low fat/etc?.

10) Decision to send is data too. Every entry should be interpreted in at least four ways: a) What the entry conveys, b) What the participant believes the entry conveys, c) why the participant chose to capture and send it, and d) what didn't they send. In short, the decision to capture an entry is, itself, data which needs to be interpreted.

That's all for now!


Share/Bookmark

Thursday, August 15, 2013

Our numbers and my favourite Jewish joke






















The time has come to share.

We are 27 people shy of having 7,000 signups on the site. And I am not including research participants.

Of these signups, 1,085 of them have started projects. Many have more than one project under their belts. Can't imagine why so many sign up but never start a project.

We have had 2,493 projects take place.

With 4,586 invited participants.

Sending a total of 39,815 entries.

Look into these figures whichever you like. But let me sum up by telling you my favourite joke. A Jewish Joke (all the very best ones are):

Old Jewish man is knocked over by a car on 5th Avenue. Crowds gather before a paramedic arrives and gently rests the old man's head on his knee.

"Sir, are you comfortable?" asks the paramedic
"I make a living" the old man replies.

Share/Bookmark