Please enable JavaScript.

Coggle requires JavaScript to display documents.

Cognition and Technology (Emotion (Persuasive tech (libertarian…

- - - - Mobile study
        
        Resource Competition Framework
        
        video from 4 cameras
        
        data coding meeting with calibration session
      - Cockpit study
        
        video/audio data
        
        transcription with action annotations
        
        data coded
        
        Theory of Distributed cognition
    - - Useful for designing new technology
      - See how ppl interact with each other
    - - naturalistic observation
      - augmented with interviews, Qs, study of protocols
    - - In the Field
        
        BIG PICTURE
        
        how ppl interact
        
        different roles
        
        general context
        
        tools used
        
        ex: Cockpit study
        
        over a long time
        
        min. day
        
        possibly weeks/months
        
        Data collection
        
        opportunistic
        
        unobtrusive
        
        plans helpful
        
        easy to forget to collect data
        
        ex: Robson
        
        Feelings
        
        Goals
        
        Time
        
        Events
        
        Acts
        
        Objects
        
        Activities
        
        Actors
        
        Space
        
        Level participation
        
        Outsider
        
        Need to supplement with interviews/reading
        
        Apprentice
      - In the Lab
        
        DETAILS OF HOW
        
        controlled tasks
        
        pro: comparing participants
        
        drawback: users not invested in tasks
        
        drawback: tasks maybe not representative
        
        controlled environment
        
        ex: Mobile study
        
        Data collection
        
        written notes
        
        audio/video/screen cap
        
        possibly eye tracking, biometric, interaction logs
        
        Consider
        
        how much time you have to analyze it
        
        what you need to know
        
        CODING
        
        decide what events you are interested in
        
        what attributes characterize these events
        
        ex: switch of device focus
        
        ex: switch "from" what and "to" what
        
        slowly go through each stream of data, marking events and extracting attributes
        
        Allows you to turn QUAL data into QUAN data
        
        Pilot study first
        
        Check for:
        
        setup effectiveness (cameras capture everything?)
        
        task descriptions clear
        
        data collection goes smoothly
        
        data supports the overall goal
        
        hypothesis promising
        
        Think Aloud
        
        Pros
        
        Insight into what ppl are THINKING
        
        insight into MOTIVATIONS
        
        Cons
        
        subconscious justification
        
        Talking changes how ppl do the task
        
        ppl don't always know why they do things
        
        can feel unnatural
        
        Alternatives
        
        Retroactive Thinkaloud
        
        Procedure
        
        ppl explain whys afterward
        
        video taken during task
        
        pro: task unaltered
        
        con: whys less reliable
        
        Constructive Interaction
        
        Procedure
        
        two users do a task
        
        talk during the process
        
        pro: more natural
        
        con: task still affected by concurrent thinkaloud
        
        Demographics
        
        Self reported
        
        gender
        
        age
        
        experience level
        
        Tested
        
        color blindness
        
        spatial ability
        
        short term memory span
        
        skill (in System I or II?)
        
        Triangulation
        
        Data from multiple sources, ex: times, ppl
        
        Multiple investigators collected and coded
        
        Interpreted using multiple theoretical frameworks
        
        Data from multiple techniques
  - - - who
        
        domain experts
        
        users
        
        stakeholders
      - reveal regulations/customs/culture
  - - - actions
      - feelings
      - context
      - social
    - - freeform
      - prompted
- - - - how we parse objects
      - how we scan scenes
      - Borders over color blocking
- - - - ex: :( new error screen VS blue screen of death / too much info
    - - anthropomorphic
        
        we are more forgiving
      - human
        
        has to be really good
      - depends on the application
        
        tools
        
        we don't want their help
        
        personal assistants
        
        we expect a helper
        
        I disagree....any app could use one, depends on how well it's done
    - - emotional affordances
      - ex: FB "end of yr" videos
      - Google doodle