Docstoc

Fieldwork and Prototyping

Document Sample
Fieldwork and Prototyping Powered By Docstoc
					         CS 376
Fieldwork and Prototyping
       Shailendra Rao
      Abhay Sukumaran
   Fieldwork / Prototyping
Fieldwork: Contextual Design
Prototyping: Prototyping for Tiny
Fingers
Case Study: Informing the Design of an
Information Management Systems with
Iterative Fieldwork
Fieldwork: Contextual Design
Ch2: Gathering Customer Data
Ch3: Principles of Contextual Inquiry
 Gathering Customer Data
Marketing vs. design
– Market: money, barriers to entry, niches
– Design: work structure, attitudes, affordances


Quantitative vs. Qualitative
Eliciting customer needs is difficult
IT people tend to focus on technology; don’t
share the user’s perspective
       Good Designers Can
Delve into intuitive processes and make the
knowledge shareable

Using Contextual Inquiry
–   In the field
–   Co-discovery of needs with user
–   Shared interpretation
–   Directed by design purpose


Marketing tells you what will sell; design tells you how
best to build it.
Principles of Contextual Inquiry 1 of 5

 The Master/ Apprentice Relationship Model
  – Design team learns about users’ work like an
    apprentice learning from a master
  – Observation + Discussion
  – Go to workplace to see work as it unfolds
  – No generalizations, but actual instances of work
  – Current events trigger past events
  – Environments and artifacts matter
Principles of Contextual Inquiry 2 of 5

Four Principles of Contextual Inquiry
  1.   Context
  2.   Partnership
  3.   Interpretation
  4.   Focus
Principles of Contextual Inquiry 3 of 5

 1. Context
 “Go to the customer’s workplace and
 see the work as it unfolds” (Whiteside and Wixon 1988)
  – Summary vs Ongoing Experience
     • Details are hard to talk about, but easy to see
  – Abstract vs. Concrete Data
     • Real artifacts and specific events make
       essential details salient
Principles of Contextual Inquiry 4 of 5

 2. Partnership
 “Make you and the customer collaborators in
 understanding his work”
  – Withdrawal and return = watching work +
    discussing how work in structured
  – Design is truly user centered when you start by
    investigating work structure, not by bringing a
    prototype
  – Other Relationship Models to avoid:
     • Interviewer/ interviewee
     • Expert/ novice
Principles of Contextual Inquiry 5 of 5

 3. Interpretation : What do these observations reveal
 about underlying structure?
  – Correct interpretations lead to effective design
  – Check the interpretation by walking the user through it
  – Be open-ended and pay close attention to nonverbal cues

 4. Focus : steer, but don’t blindfold
  – Follow up to probe for detail on relevant things
  – Use intrapersonal triggers to expand focus
  CSI: Contextual Inquiry
Crime Scene Investigation… the
Contextual Inquiry season
How does Kumar manage his music
collection? The Crime Scene
Investigator… commits a crime!
Roles
– Shailo: Kumar the customer
– Abhay: Harold the interviewer
             Crime 1
S: Hey I’m Kumar and I’m here for the
study. Aw man, it was really tough
finding the place! I took 101 and then
the Dumbarton bridge and then 880
North and then got stuck in traff-

Not in users context
             Crime 2
S: {feeling awkward} So, Harold how
about them Warriors this upcoming
season? They look good man!
A: Um, I’m not really interested in
basketball. I’m more of a baseball fan.

No sensitivity to culture and not
making the user feel comfortable.
             Crime 3
A: Okay… moving on. Have you ever
converted a CD to mp3 format with
your computer?
S: Yup.
A: What program do you use to do this?
S: Well I use iTunes.

Leading question
                                     Crime 4
A: So today I’d like you to tell me about how you manage your music collection. Do you listen to mp3’s on your
computer?
S: Yes.
A: What program do you use to manage your music?
S: Hmmm, I think it’s called iTunes. That Apple one.
A: Have you ever created an Audio CD from your MP3s?
S: You mean to listen in my car?
A: Yes, whatever.
S: Oh yeah I’ve done that plenty of times.
A: How many songs do you have?
S: You mean mp3’s or cd’s?
A: In iTunes.
S: Probably 20GB
A: So how many songs is that?
S: Hmmm, a lot?
A: Okay… moving on. Have you ever converted a CD to mp3 format with your computer?
S: Yup.
A: What program do you use to do this?
S: Well I use iTunes.


Interviewer/ interviewee relationship
model
                      Crime 5
A: Well actually you know Nero does a better job of
converting CD’s to mp3 format. It’s much faster. You should
use it.
S: Oh yeah? You know I’m probably not as well-versed as you
in this computer music thing.
A: Well let’s just say I’ve been doing this for years. I’ve been
converting CDs to mp3’s since 1996.
S: Wow, Harold! I didn’t even know mp3’s existed back then!
A: You have so much to learn, Kumar.


Expert/ Novice relationship instead of
Master/ Apprentice
                Crime 6
A: OK. So now it’s exactly as if you were
setting at your desktop computer with your
normal program and settings right?
S: I guess so. This is actually my friend’s
laptop that I borrowed just for this study.
A: Well okay, but it’s exactly the same right?
S: Sure. It’s a computer with music right?


No Context- studying an artificial situation
and setup
            Crime 7
S: Pretty good.
A: What do you mean? Tell me three of
the biggest problems you have had.

Leading, assuming that there have
been problems
             Crime 8
Uh, I guess sometimes a few of my files
don’t transfer over.
A: Because the Bluetooth broke down?

Leading the interviewee
          Crimes 9 & 10
S: Um, no, I don’t think so. Is Bluetooth the
wire that connects the Ipod to the computer?
A: <haha>, are you kidding? No, it’s a short-
range radio frequency standard for mobile
device communication.

Expert / novice relationship model
Probe for concrete data vs. abstract
            Crime 11
S: ok. Here goes. <creates two
playlists>
A: <watches > Ok. Did you have any
problems doing that?
S: No, I don’t think so.

Interviewer did not alternate between
watching and probing.
               Crime 12
A: I think you were trying to replicate a saved
query there.
S: uh.. yeah, the thing is, I don’t know how to
browse my music through iTunes. I just know
how to search.

Interviewer should share interpretation
with user, and let them fine–tune it.
               Crime 13
S: come to think of it, I have a tough time
browsing the stuff I’ve TIvoed. I think the
problem is that I collect so many shows in
such small periods of time. I’ll TiVo 5 shows a
day. Actually I….. Do you ever have that
problem, Harold?
A: um, no
S: Well, that’s pretty cool …

Focus was not handled properly
             Crime 14
A: Yeah, next time I’ll show you how to
manage your music much more
effectively.

Reinforcing the wrong relationship
model (expert/novice)
Prototyping for Tiny Fingers


Fudd’s first law of creativity: “To get a
good idea, get lots of ideas.”
Lo-fi (Paper) vs Hi-fi Prototyping
“Know your user, you aren’t your user”
  Not so Tiny Tradeoffs



What are the Tradeoffs to Lo-Fi
          Prototyping?
     Advantages to Lo-Fi
        Prototyping
Quick to build (especially multiple)
Get user feedback fast
Keeps focus on conceptual elements rather
“Fit and Finish”
     Advantages to Lo-Fi
        Prototyping
No false impressions of how much backend
work has been completed
Avoid debugging
Great for choosing between different several
different mockups
  Disadvantages to Paper
       Prototyping
Can’t sell it
Bad for testing look and feel
Can’t show a detailed proof of concept
Can’t test changes to an existing
system
Could encourage excessive focus on
micro-elements
           Lo-Fi Quickies
Think back to preschool-- Get your hands
dirty
Expect the unexpected from users-- Practice
for various actions
Stick to your roles (Observer, Wizard of Oz,
Facilitator**, and User)
– Only Facilitator should be audible and “visible”
Use realistic scenarios
Use domain relevant sample data
           Case Study
Informing the Design of an Information
Management Systems with Iterative
Fieldwork
               PIM design
Key question: Why does paper-based PIM
persist even in the face of advantages of
online formats:
–   Searchable
–   Shareable
–   Easily archivable
–   Auto-generated reminders

How to integrate paper into the new PIM
Or – how to simulate affordances of paper
      Paper affordances
Permanent
Lightweight
Form factor – can fold, tear, etc.
Universally available
Supports grouping (piles on desk),
association (paper clips), copying.
Anything else?
  Pilot Interviews (Phase 1)
Reality check
– proposing a radical idea (paper PIM)
– fish around and get a sense of the domain
Ask:
– How do you do it now?
– What’s wrong with it?
– Could our design make things better?

Outcomes:
– Tangible, context-embedded reminders
– Temporary clumps of documents (e.g., all emails+docs for a
  meeting)
In-depth interviews (Phase 2)
Explore evolving ideas
Ask:
 – How and why do people organize on paper and offline?
 – Physical artifacts & their organization
 – Could we print and scan?
Outcome
 –   Filing is difficult
 –   Scanning isn’t going to work
 –   Sticky note reminders are great
 –   Documents of different types need to be grouped
 –   All sorts of different tool combinations used
In-breadth interviews (Phase 3)

Back up and distil findings
Ask:
– Email usage
– Organizing meetings
– Taking notes
Outcome
– Design to embed in everyday apps
– Simple, flexible
– Something like sticky notes to label / group
      What they did right
Didn’t just look for confirming evidence
of their initial hypothesis
Paper prototype – no software to get
attached to
In – context, iterative user research
Understood importance of email for
coordination and collaboration
        What they did wrong
No observation of users
Diary study might have been effective
Scanner oriented gadget bubble
Ask them to project “whether it would work”
Slicky does not equal sticky
Making a customizable UI does not mean that people will
actually take the trouble to customize it
Problems with groups – people don’t file because they have to
think hard about categorization.
They do needs finding in the user domain, but the solution is still
from an engineering perspective
“Raton Laveur”
     Advice for the field



Go forth and prototype!

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:3
posted:3/22/2012
language:
pages:39