Thursday 12-2PM, 04-05-2007

From CENS Urban Sensing

Urban Sensing Meeting 4/05/07 12-2pm

Visit from the IRB

• Sound level campaign: recording sound levels and location -> IRB says it doesn’t need their approval since the purpose is to learn more about sound, not humans
  • Might have data protection concerns, which may need another office’s approval, but not IRB’s.
  • Outside public noise: anybody can record it w/o approval, not private space
  • Secondary inferences from sound data
    • Limit demographics to reduce risks of identity being inferred
  • If instead, experiment was to evaluate a person’s exposure to say, pollution, during a 3-hour commute, it would be considered HSR. (about humans, and recorded at regular intervals, out of person’s control)
  • Can collect exact same data, it just depends on the intent of the research (very fine line)
• Diet Sense
  • People nearby should be informed of possible exposure by camera
  • Okay as long as no faces recorded, or just blur faces out

Other info:

• Protocols determined on case-by-case basis
• Look at the definitions of research, human subjects, and human subjects research (HSR)
  • A human subject is a living individual about whom an investigator obtains either:
    • 1) Data through intervention (e.g. taking blood sample) or interaction (e.g. interview) with the individual; or
    • 2) Identifiable private information
  • Human subject research includes any activity with a human subject that is designed to produce generalizable knowledge and involves the systematic gathering of information. This research may include direct interaction or intervention with subjects or it may involve indirect activities like the analysis of human subject specimen or data.
• Look at purpose of research, whether it’s about the people, or something else like advancing technology
• private vs. public space
  • subjective
  • a sort of continuum
  • has everybody there agreed that particular space is private/public?
• Theses/publications that includes private info require IRB approval
• If recruiting members from public for pilot, then HSR, requires IRB approval
• If using UCLA resources (computers, staff, etc), it’s considered UCLA research
  • Talk to (Legal?) Council: Liability of someone misusing a university resources
• If in our pilot stages where data is used only within the group, never for outside or generalizable knowledge, not HSR
• A documentary is not research, unless it is analyzed for research
• If doing research and have some recordings of illegal activity, would we be required to testify in court?  Yes, unless have a Certificate of Confidentiality.
  • To get approval for Certificate of Confidentiality, must have a very valid reason for HSR
  • A Certificate of Confidentiality protects the investigator from being subpoenaed to court
• Can always email/submit to IRB for official determination
• Ultimately, it seems that the IRB is not our tighter bound on campaigns (not about data sharing protection)

Discussion about Spring Quarter

• Dealing with Find?
• Focus on 3 campaigns

1. Sound Level campaign: Elisa, Stephen (automatic agreement since he’s absent); also prepare for Summer campaigns
2. Diet Sense
3. Sidewalk: Sasank, Vids?

• intentional image capture
• feedback to users, key places to calibrate how they’ve done or where else to go

• have 1-2 people focus on these campaigns for the quarter, even if it’s just to get people to do these campaigns
• let’s have definite goals for each of these campaigns: i.e. for sidewalk campaign, intentional image capture
• also note: we’re not absolutely set in those goals, they’re flexible, if they change with your interests, that’s okay
• have someone go to Sensorbase mtgs about images & ? (Gong, Nathan, Deborah, Mark)
  • if ever a problem with SB, email them and cc Mark, Deborah, Jeff

Action Items

• Elisa/Stephen: present Sound Level Campaign results at Week 3 meeting
• Vids: send out script to extract all of your images from Sensorbase.
• Everyone: use Vids’s code, go through and delete out pics of non-food, keeping only good, full pic, non-blurry images for now. Save as zip file and bring to next meeting.
• Deborah/Jeff: Email ___ when all good images are on Flickr.