CS631p Human Computer Interaction
Spring 2001
Lecture 3

Using Video to Prototype User Interfaces
Laurie Vertelney

How to decide what interfaces to build:

  1. Specify what tasks the interface will enable.
Begin with:
    1. User Interface Engineering Requirements Specification
    2. Lists of Functions and Commands
    3. Scenarios or Stories
  1. Design how users will interact with technology to accomplish task.
How do you prototype user interface ideas?

Apply to Interface Design via Prototypes

  1. Animated Prototypes
    1. Animated Drawing
      1. Make sketch of interface
      2. Photocopy sketch to make multiple copies
      3. Draw atop copies to show interaction events as in-betweens
      4. Place drawings in sequential order under video camera (shoot 1 drawing for 10 to 20 sec.)
      5. Edit video.
    2. Cutout Animation
      1. Objects cut from paper or cardboard and place atop background
      2. Film as they move under video camera
      3. Edit video
    3. Animated Objects
      1. Three dimensional objects
      2. Move under camera with magnets or fishing line
    4. Computer Animation
      1. Cell Animation
  1. Computer Scripting Interactive Prototypes
    1. Hypercard
    2. Shockwave
  1. Mixed Media
Advantages:


Disadvantages:

  • - Expensive
  • - Difficult to simulate
  • - Not Interactive


    • Sample Storyboard Software:

    Methodology Matters: Doing Research in the Behavioral and Social Sciences

    Joseph McGraph

    Basic Features of Research Process

    1. Some context that is of interest
    2. Some ideas that give meaning to that context
    3. Some techniques or procedures by means of which those ideas and contexts can be studied
    Research deals with several levels of phenomena: relations between elements within a context

    Three domains:

        1. Techniques for controlling
          1. Experimental control
          2. Statistical control
          3. Distributing the impact (Randomization)
      1. Relations - applications of various comparison techniques
        1. Used to assess relations among two or more features of human system
        2. Comparisons involve three sets of features of the systems under study:
        1. Comparisons assess co-variables or association between independent and dependent variables
    Research methods as opportunities and limitations
    Dual nature of methods: All methods flawed - must minimize flaws

    Summary:

    Fundamental principle in behavioral or social sciences:
    Credible empirical knowledge requires consistency or convergence of evidence across studies based on different methods

    Research strategies: Choosing a setting for a study
    Research evidence - somebody doing something in some setting

    What we can ask:

    Maximize three criteria:
      1. Generalizability of evidence over actors
      2. Precision of measurement of actions
      3. Realism of context
    Can not maximize all three simultaneously - kind of like the uncertainty principle

    Increasing one reduces other:

    e.g. increase precision - extending control of variables reduces realism
    e.g. field study - increases naturalness reduces precision
    Study Design, Comparison Techniques, and Validity
      1. Data collection
      2. Aggregation and partition
      3. Comparisons made
    Comparison techniques: Assessing associations and differences
    Randomization and "True Experiments"
    Randomly assign cases to conditions
    Sampling, Allocation, and Statistical Inference
      1. Statistical analysis (e.g. coin toss)
      2. Probability
    Validity of findings
      1. Internal validity - How close you can say that X varied Y - must be able to rule out other plausible hypotheses
      2. External validity - How well your findings will hold-up upon replication
      3. Threats to validity - plausible rival hypotheses
      4. Construct validity - How good is theory
    Classes of Measures and Manipulation Techniques
    1. Kinds of data collection methods:
      1. Self reports or questionnaires
      2. Observations (visible vs. hidden observer)
      3. Archival records - material already exists
      4. Trace measures - behavior itself leaves trace
    (All have strengths and weaknesses)
    1. Techniques for manipulating variables
      1. Selecting cases with desired values and allocating them to appropriate conditions of study
      2. Direct intervention in systems under study
      3. Indirect inclusion of variables
    Three approaches to manipulation:
      1. Selection - directly select a choice
      2. Direct Intervention - set up test cases
      3. Inductions

        4.  - Misleading instruction
         - False feedback
         - Experimental confederates
         
         
    Usability Inspection Methods
    Mack & Nielsen

    Usability Inspection - evaluators examine usability of related aspects of UI

    Characteristic - reliance on judgment as source of evaluator's feedback Four ways of evaluating UI:
      1. Automatically (run through software)
      2. Empirically (testing interface on real people)
      3. Formally (using exact models or formulas)
      4. Informally (based on rules-of-thumb) <- Usability Inspection
    Best testing: empirical and informal Inspection Objectives

    Inspection occurs when design has been created and utility toward users needs to be evaluated
    Concerned with classifying and counting number of usability problems.
    What is a problem? - any change in design that leads to improved system measures

    Usability engineering lifecycle - problems - design team must redesign UI which means further analysis, design, costing

    Inspection Methods

    1. Heuristic evaluation - most informal - specialists judge whether each dialogue conforms to principles
    2. Guidelines reviews - cross between heuristic evaluation and standards inspection
    3. Pluralistic walkthroughs - meetings where users, developers, and human factors people step through a scenario
    4. Consistency inspections - designer representing multiple projects inspect an interface
    5. Standards inspections
    6. Cognitive walkthroughs - detailed procedure to simulate a user's problem solving process
    7. Formal usability inspection - like formal code inspection (moderator, scribe, inspectors, etc)
    8. Feature inspection - looking for functions