Selecting the Invisible User Interface Development Tool

1
Selecting the Invisible User Interface
Development Tool
Joanna Lumsden Glasgow Interactive Systems
Group Department of Computing Science University
of Glasgow jo_at_dcs.gla.ac.uk www.dcs.gla.ac.uk/jo

2
A difficult but important decision

• Influence
• UIDT can influence 50 - 70 of application code
• Benefits
• iterative design, help undo, reduced generation
  cost, range of specialists, UI consistency
• Acceptance
• support for working practice
• support for application requirements
• resonance with users (learning curve)

3
The selection decision

• Selection Decision
• identify assumptions restrictions
• represent problem to facilitate solution
• focus information gathering
• decide between alternatives
• Aim for Good Quality Decision
• Importance of Context of Use
• people functional requirements
• support full effective use of contextual
  information
• make relationships between context problem
  explicit

4
SUIT

• Framework/Reference Model
• extensible
• environmental context of use, UI requirements,
  developmental context of use
• Context Sensitive
• tailoring based on context of use
• 2 Step Process - Multiple Uses

UIDTs
5
Routes through SUIT
6
Routes through SUIT - A
7
Routes through SUIT - B
8
Routes through SUIT - C
9
Empirical Evaluation

• Aims
• can people use SUIT?
• Setting
• 27 undergraduate students
• evaluate 2 web-page authoring tools
• Results
• Q is it possible to do an evaluation using SUIT?
• A yes, but at high costs (time effort)
• Q how did method affect subjects perspective?
• A focussed attention on specifics of context of
  use

10
SUIT Visualisation Environment
11
Case Studies...

• Aims
• investigate how visualisation environment is used
• to create initial model(s) of analysis
  strategies
• Setting
• 5 experts context sensitive evaluation of 2
  UIDTs
• video-taped direct observation, think aloud
  protocols, interviews
• content logging (verified) verbatim
  transcripts
• Analysis
• usability analysis
• content logs examined for action sequences
• produce observational model of analysis
  strategies

12
Modelling analysis strategies
13
Expanding the SUIT Route Map
14
In summary...

• SUIT shown to
• be viable ?
• focus attention on context of use ?
• have potential to tackle trustworthiness ?
• audit trails, peer review, etc
• be compatible with the Capability Maturity Model
  ?
• Accrued
• useful and practicable visualisation environment
  ?
• knowledge of UIDT data comparison strategies ?
• modelling notation for comparison/analysis
  strategies ?
• So where now?
• expansion into other fields of evaluative
  selection ?
• further evaluation ?
• open-source community of SUIT users ?

15
Selecting the Invisible User Interface
Development Tool
Joanna Lumsden Glasgow Interactive Systems
Group Department of Computing Science University
of Glasgow jo_at_dcs.gla.ac.uk www.dcs.gla.ac.uk/jo