Bay Area July 2007

hide

July 11, 2007, 6pm
Eszter Hargittai
Topic TBD - Eszter's recent work in online use

Socialtext/coworking office at High and Forest,Palo Alto

Upcoming invite: http://upcoming.yahoo.com/event/217711/?ps=6

Participants

Notes on July Wiki Wednesday:

Eszter describes the web use project, and some background for the designs of its studies. Its site is: http://www.webuse.org/ Some of this work has been funded by the MacArthur Foundation.

People's view of the digital divide is too simplistic. What is people's use of IT after they go online? Also, what are the real generational differences rather than the mythical ones - do they match? (The myth: Gen Y juggling, Gen X in middle, Boomers a bit incompetent). To what extent does the net enable social mobility, or reproduce existing social inequalities?

A nifty chart of the user, socioeconomic context, their technical and social contexts, how it feeds into their skills (measured in "Eleven Dimensions of Skill") which then informs their types of uses of the Internet. We digress briefly into ideas on enhancing capital — human capital, social capital. How do these skills and uses then feed back into the user's social class? (Implied: how can we hack that feedback loop to enable social mobility and... I would go so far as to say, social justice. Specifically, where in education, and in software design, can we work for change? )

Adina's question - do you look at "skill improvement" as information seeking or what about complicated social things like, using flickr improving your photography skills through social feedback?

Eszter: Showing relationships but not causality. Need longitudinal data. Examine their photography skill now and then over time. Then control for things like propensity to learn skills from other factors. So it is complicated.

One study measured total time spent on eight tasks. A cumulative distribution function. How long the person took to find the information - tax forms, health info, job search, political info. 43% failed on political task. (in 2001). Find a third party web site that compares presidential candidates' views on abortion. Given unlimited amounts of time, people couldn't do it. This has implications for people's use of the net for political education.

Age is the only predictor when you take out regressions, control for experience using the internet, time online per week, years exp. with computers, etc.

Even at an elite school with privileged students the incoming freshmen do not understand online skills... Am doing this study at University of Illinois Chicago. Urban campus, non flagship of its system, 10% ethnic diversity among research universities. Required course on campus - 85 sections, one course. paper/pencil surveys because you cannot assume skill to take a survey on the computer. this controls for that. it's a 20 page survey and must be entered manually by interns... 44/56% male/female

Questions that serve as proxies for measuring skill.

Basic internet and web 2.0 related terms:
adv. search
bcc
blog
bookmark
bookmarklet
cache
favorites
firewall
frames
jpg
malware
newsgroup
pdf
phishing
podcasting
pref.setting
rss
social bookmarking
spyware
tabbed browsing
tagging
torrent
web feeds
weblog
widget
wiki

There were also some madeup terms , very funny! But people were not fooled on the whole. Earlier studies measured very basic terms like "email" and "browser" but the variance was very small even with different age groups.

Important factors that affect skilll:

  • women - hispanic origin - parents level of education
  • autonomy: higher # of access locations meant higher skill, # use locations
  • experience - use years , weekly web hours

user of services that empower users
we are placing bets on what is used by these students most. usage.
my predictions

  • youtube
  • myspace
  • wikipedia

eszter's results:

  • wikipedia 85%
  • youtube 81%
  • facebook 79%
  • second life 1%

Eszter confirms some of danah boyd's thoughts about socioeconomics and facebook vs. myspace. A strong correlation with parental education level. We do not have enough data or funding for the proper sorts of studies yet.

There was a variable for all of delicious, reddit, digg, does it correlate with their checking off that they understand what social bookmarking is? Yes. So if a person said they knew what social bookmarking is, they are highly likely to be a user of a social bookmarking site.

User background and types of activities.

Types of uses of the net.
Women use the internet for fewer things than men do.
Hispanic origin
lower levels of education parents

Adina mentions the carnegie mellon study about programming. age at which you are exposed to computers.

Eszter explains these are regression studies, not simple correlations.

Interesting new dimension to research looking at measuring skill before they reach college and then when they are sophomores, looking at roomate assignment.

Multitasking study with funding from the NSF. data collection on text messaging. and they text back the researchers every hour with what they are doing.

Most of people's assertions about things like "young people use IM, old people use email" are not backed up by data. The studies have not been funded.

another study on gender differences in creative media - private or public posting of photos, video, etc?

A break and some quick talks and demos.

Newton is showing off his mashup of google maps and upcoming. You can search for all events on upcoming within x miles of wherever you click on the map. You can also narrow the upcoming search by tag and category and other criteria. The flickr / google maps mashup was very interesting. we could see yesterday's baseball game very clearly on the map of san francisco — so clearly that the shape of the stadium was visible, and we could look at the satellite photo to see the stadium itself and who was there.

Gordon made a quick hack to use socialcalc 1.0 and scrape the data and then display it and you can sort and filter the data and stuff.

Tim Bonneman is talking briefly about online dialogue and deliberation and using wiki-based approaches.

activity: low edits, small numbers of users. Not getting better over the last few months.

is this a wiki problem?
or that the sites aren't getting good enough press?
ideas for further research?

Adina recommends looking at:

  • fluwikie
  • dkosapedia

both come from established communities

Eszter - doesn't matter if it's a blackboard, mailing list, blog, etc. tiny percent of people participate on any of them. That, we have research into.

Tim answers that he wants to think about what tools could help facilitate deliberation and dialogue -

Eszter mentions James Fishkin from Stanford and Tim already knows him from NCDD.

Luke's video for Vancouver Wiki Wednesday on their bug tracker wiki hack.
http://one.revver.com/watch/328515 And the Vancouver work from home club.

Steve Bang is talking about collaborative writing. Adoption problems. Why don't people contribute? What are the barriers to adoption and use? Was at IBM, now at SAP. nobody in his new group knew if there was a wiki for his group or not. talking to people and then they got productive and showing them how to add things. Initial apprehension and need a personal push to start. something about the design of all wikis, confluence, socialtext, mediawiki. what is it?

pete: not design as much as it is a new tool. any new tool. they need to see someone else using it. email adoption worked that way when email was new. a friend shows you how cool it is.

Steve: When people are working remotely... with distance, how can you show them?

Pete: talk them through it.

Steve's group ... here's the 5 basic documents. how to create a page. a link. etc.

Liz: screen casts are good

Pete: but the step before that is why do they need a wiki? how will it be useful to them?

Steve: but they know basically .

Adina: the big blue edit button, cheerful cartoony friendly

Pete: adoption of adding stuff that people want to see a lot like lunch menu and phone numbers. put in one or two people's phone numbers wrong.

Adina: ward cunningham trick. put in a few small mistakes so that people go in and fix it.

Liz: deliberate gnome errors so that a wiki gnome will go in and fix it, maybe starting as a gnome is an easy entry method

Adina: add your agenda item and add a link to your particular thing. notes and action items are right afterwards.

Steve: confluence bar, tabs, wrong or right? tabs across, the different layouts. one the edit at the top, one it is in the nav bar, 3 people said where is the edit. "page operations" Even "edit" they think of as what someone does to my writing , does not mean "contribute" to people.

Gordon is demoing a use of pageoftext.com.

Now links are in pageoftext.com. So now you have search! on pageoftext.com/link_demo . trees that work out from there. So you can see a link tree which also functions as a recent changes. now it has an rss feed too.

unlike workspace where you have different workspaces... you can have any number of trees and a new root node that shares pages . so from one root node you can see whatever it links to. but from another you cannot find the parent node.

Liz: so it is a different method of privacy.
Adina: or obscurity
Liz: so it abandons static workspaces and the same pages are viewable in different structures.

Thanks to everyone for coming! Though there were only a few of us at the event, we had some great conversations.


Click this button to save this page to your computer for offline use. Created by Tim Bonnemann on Jul 4 4:31pm. Updated by Harry Wood on Feb 2 9:53am. (28 revisions, 8,221 views)