SESSION 2

2B BARRIERS

Easiness to use -- Uploading - Privacy concerns -- ACL -- Footprint -- "Combination" of controlling functionality is one of the key to think about barriers. -- "Easy to use" is complicated because of it's not single-function issue. If we want to share photos, everything on the process should be easy, from taking photo itself, send or upload photos to the sharing server, controlling ACL for the photo. To use some kind of *Wizard" to execute these whole steps may be a solution to lower the barrier.

At this stage, it seems difficult for the users to share pictures via web or MMS. Many people feel that photos taken are more personal/private, and are sharing mostly with a restricted number of people. Many of us keep photos inside the terminal, and share them face-to-face. Cost of sending photos may not be a barrier; presumably, people care more about psychological cost of sending and sharing them.

A recent “footprints” feature on SNS may be a new way to control access to one’s personal space.

Picture sharing supports face to face communication. This is currently done predominantly by showing phone screens to each other. It seems that simular face to face features are called for when sharing photos in information networks. An example of this is passwords to ensure that the sharing is limited to small, specified groups.

Social technologies such as footprints could provide metadata making sharing more interesting.

Opening question: Is PICS something that people actually want to do?
- what's behind the diversity in opinions?
- is there really a trend, or is this just a spike in early adopters?

Risto: producers vs consumers, about a 1-to-10 ratio in practice.

Maybe people forget that they *can* do mobile-to-mobile, and that's
why it isn't totally pervasive..

Video conferencing was supposed to be a big thing.. but never really
caught on. Is MMS the new videophone - full of promise, but not going
to really deliver?
- maybe photoblogging is a technology that will catch on.
- Q: why would this be different? what is the structural/social
factor(s) that would make it different.

Group shared photo / info spaces.. Questions include:
- does it lead to an expectation that people will *always* be up to
date? does this cause problems when individuals are not keeping
up to date?
- are there examples where groups *do* keep up with a group of
others in on online space successfully?

Bloggers: in the beginning there were a few famous bloggers that
showed the way for everyone else..
- do public picture sharing sites have "famous users" that are
showing the way in this area?
- the Bento-Blog, perhaps that's an example

Young people in Japan how often do they use phones to share photos?
- friend-to-friend, about once a week.
- more often into moblog..

Norio: Japanese weblogging site allows levels of access to one's photos:
- Friends / friends-of-friends / etc..
- people tend to share more, since they have ways to control access.

Question: is there a trend towards smaller groups, private groups?
- families, close friends, intimate groups...
- blog-people were surprised about this, since the original idea for
blogs was to allow one-to-MANY publishing.
- Fumi shares that sensibility, and adds that people want to see
who is accessing your page. (recent social networking systems
can leave "footprints" - you can browse who has accessed your
page)

Interesting ideas about footprints:
- can facilitate the creation of common ground, when you know who has
read your blog you don't have to ask/make sure.
- when we browse, do we want to have the option of *not* leaving
footprints?
- "invisible mode"
- "anonymous mode"
- "erase my tracks" command, etc..
- yes, we want some ability to control how/when we leave "footprints"
but we are not totally in agreement about the form that this takes.
- perhaps footprints are an alternative to passwords on a blog - if
I am a content creator, maybe I want to enforce non-anonymity,
rather than specifically restricting access to my content to certain
people.
- having foorprint traces can be a useful tool for law enforcement,
when necessary. If the authorities can look at patterns of footprints
then they can catch "creepy" behavior potentially.
 

We had a wide-ranging discussion of different issues influencing adoption and uptake and what kind of features might contribute to adoption by different groups.

There was skepticism that MMS is ever going to be a pervasive model of image sharing. MMS is probably going the way of video conferencing – it is not the right set of technology tools to support the social practices of communication and image sharing. By contrast, there seemed to be agreement that posting to picture sites of blogs is a growing and emerging practice. Among the Japanese students in the group they described how sending an image over MMS is relatively rare, but about half the students at the Keio SFC campus have a picture blog of some kind.

Cost, at least in Japan and Finland did not seem to be the major barrier to sending MMS. It was more an issue of social appropriateness. People prefer to share off their handheld screens versus sending. If the person is not physically nearby though, there may be some extra motivation to send an MMS.

Technology ease and appropriate system design seems clearly to be one factor hampering adoption. Combined with the lack of existing social models for pervasive picture sharing has led to relatively slow adoption in many areas. By contrast, pervasive picture capture seems to be more intuitive to users, and fits within existing interactional modalities of personal archiving.

One issue hampering adoption seems to be the uncertainty about privacy and audience. One system described allows for easing setting of groups and access, which seems to encourage sharing. We also discussed the idea of enabling “footprints” so people can see who has been accessing their site. This would allow for some kind of barrier to access short of password protection or clear group definition. Privacy concerns of opting in or out of leaving footprints would need to be resolved.

 

Is pervasive photography still in the novelty factor stage? It is probably in different stages in different parts of the world, for example, within Japanese young it might have already found its normal use. In Europe and the US still has not found the large masses even within demogrpahic groups.

For other than nearly adopters, there is the issue of learning the technology and conceptualizing all the new ways of sharing the photo. The pervasive capture has probably taken of because the learning curve is less steep than in pervasive sharing, where the different options (sharing face-to-face, vie mobile network, via PC), privacy management etc. come into play.

Will the cameraphone (pervasive camera) replace the regular camera after all? It sounds from examples from the Japanese that each camera has found its own use (small camera photos for mobile sharing, and larger photos for other purposes).

When you show a photo from a screen, you do not give ownership of the picture to that person. How much this affects use? This is probably dependent on the privacy management provided.

People have more control over text as a media than the control over photos. Text is 100% created by the person herslef but with photos the level of manipulation and editing is limited.


-is this PICS a novelty, will it go down?
-mob 2 web, mob 2 mob
-people do not remembr that hey can share
-sharing face-to-face, via mobile network, plugging to PC
-blog
-pervasive capture is happening, but is pervasive sharing not happening. Using only for novelty.
-picture-sharing 5yrs in Japan (P2p once a week & community mo-blog). Cam phone picture smaller than big picture, easier to send & consume. What if camphone founds its slot and does not replace the digital camera.
-showing picture from screen, you don't give ownership of photo.
-control of text vs. control of photo, cannot control the content of photo. Control of audience.
-50% of "Japanese friends" have sites with photos


 

We talked about the barriers of sharing photos and barriers. A “Footprints” metaphor was interesting. Knowing who is watching moblog is key to the privacy and sharing. Especially talking about group picture sharing, footprints will change the social practice of sending receiving sharing pictures.

 

2C SOCIAL SCIENCE METHODS

 - I don’t like methodology as a particular problem – unless we stick to a few methods & start to screen out research based on methods
- Surveys are not a problem, they do not tell much about camera phones anyway
- What kinds of statistics to use in large-scale studies? There are lots of statistical tools to “condense” observations that can then cue qualitative studies
- How could we study “the autobiographical impulse” using empirical data?
- How to “instrument” our studies?
 

The main question considered was Marc's: how to balance between
quantitative and qualitative. It was generally recognized that (1)
they answered different sorts of questions and that (2) quantitative
(cheap) was good for cueing qualitative (expensive) but there wasn't
much useful agreement past that.

With regard to this specific question, I thought the main issue - what
are people trying to accomplish - was not really met head-on. Some
people were trying to understand individual practice, others trying to
inform design and individual motivations, others to understand group
social dynamics. This seemed to make the discussion go off-focus.
 

How do we find the right mix of qualitative and quantitative methods for large scale studies and for cross-cultural studies?

Issue of inaccurate or incomplete recollections by human subjects of their behavior (who they share with, where and when they take photos, etc.).

What do we mean by qualitative research; what different understandings and methods do we have:

“Cultural probes” in which various materials are used by users in lived contexts to document design ideas about future use scenarios and also current motivations.

Classic approach is to alternate qualitative and quantitative methods to help understand what to ask and what to track.  Quantitative methods can help us model behaviors; qualitative methods can help us understand interpretation and intention.

Quantitative research can help us identify user types to focus on in qualitative research.

Model of mobile media research:

• Massively instrument many users’ information flows and content
• Analyze data to see patterns and anomalies
• Use this analysis to determine what types and instances of people to pursue qualitatively
• Interview these chosen subjects
• Iterate by improving the instrumentation methods

Using cameraphones to collect data for a diary study. 

How do we determine which demographic factors are salient for cameraphone use? Traditional notions of gender, age, ethnicity, income, etc. may be less salient than the architecture of spatial/temporal/social contexts and transactions, for example:

• People who have a regularized context of socializing vs. more dynamic contexts
• People who have a need to regularly with remote (i.e., communicable only through network transmission of data) individuals and/or groups
• People who have a highly particularized individual/group/event they regularly document and share (like people who photograph their children)

We can use large scale data tracking and analysis to identify common and uncommon patterns of use, and then use that analysis to develop a user typology and/or personas that come from the observable patterns in the data.

For the past few years, a “place” of social science perspective/method has been gradually changing (in Japan).  For instance, we are getting opportunities to work with a group of computer scientists/designers to conduct a series of qualitative research including fieldwork, focus group interviews, etc.  At this point, we are not sure about to what extent we can speak about system/design issues, yet.

As for interdisciplinary studies, social science perspective tends to describe what happened, rather than what is going to happen.

 Methods (quantitative and qualitative) for very large groups?
How to find the right mix?

What do we mean with qualitative methods?
What about cultural probes? More a method for inspiration than studying people.

Methods for mobile technology?

Voice mail as a way of collecting data in the US.

Quantitative data for seeing patterns in usage (e.g. power users and users functioning as hubs). Danger with quantitative questions when simply asking questions like nationality and sex without knowing why you are asking. You need to ask if a question is relevant. For example: are there children in the household? Is gender important to know and if yes, why? How do we know which demographic questions are relevant for a study about cameraphones?

Classic approach is to alternate quantitative and qualitative methods.

How do you instrument people?

Looking at content (of e.g. images) and flows instead of text (interviews). (?)

Instrumentation:
How many pictures? To whom are they sent? Instrument the devices instead of humans…
Instrumentation of large-scale behaviour to see whom you should talk to later. Instrument as much as possible (and invade privacy).

The use of images as a trace mechanism? Images to collect data.

What is your question? Crucial to know what you are asking for when conducting a study…
 

 Perhaps because I have very limited experience with social science methodology, I found this discussion a bit difficult to follow. But what became most pronounced to me (as I, admittedly, filtered the discussion through my own scholarly interests—and dilemmas) is the difficulty that mobility/mobilities (of images, people, etc.) seemingly pose for study of camera phone imaging and sharing. For me the question becomes: how does one study multiple mobilities simultaneously? Does tracking people’s information flows (e.g., the number of images they take, when; where images are sent, when and how, etc.) provide a means for studying mobilities? If tracking information flows is not new (i.e., its having been a means for determining particular patterns of practice/use, as well as anomalies), then how must tracking be deployed differently given the variety of mobilities (across such a variety of domains/contexts) to be considered?


The creation of sociotechnical design processes relies on people working together and not just transmitting documents.

An attitude friendly to “trying things out” helps in the interdisciplinary collaboration.

Problem of trying to publish this work which does not fit in a clear disciplinary area.

Read Lucy Suchman’s article on “Located Accountability” about social science and engineering collaboration.

Interdisciplinary collaborations need people and mediating artifacts (especially diagrammatic representations) working together.

Mimi says that age is a much more important determinant of technology adoption and use than geography (especially of nations).

National and cultural categories tend to be a knee jerk assumption for explaining differences in behavior, which may actually be less important than age, technology access, urban infrastructure, etc.