Task Analysis

Task Analysis

• Analyzing and describing how people do their jobs/work

   --> Go to their environment 

• Examine users’ tasks to better

Components 

       understand what they need from interface and how they will use it.

• Three key components to include in discussing how people work
   - Activities
   - Artifacts
   - Relations
• Don’t just focus on computer system artifacts and interactions
• Study related processes and  objects in the environment that people may use and involve
   - Example: office env---papers, whiteboards, etc.

Task Analysis Focus

• Focus on observable behaviors
   - What are the practices, methods, steps, objects, …, used?
• Observe users, what they do, less so how they do it
• Not on internal cognitive state of user (more on that next week)

Input & Output

• Gather data:
   – Documentation
   – Interviews
   – Observation
   – Surveys/questionnaires
   – Automatic data recording/tracking
• Represent Data:
   – Lists, outlines, matrices
   – Narratives
   – Hierarchies & Networks
   – Flow charts

Data to be Gathered

• Information about users
• Description of environment
   – Where the tasks will be performed
• Major goals of the job
   – What will result in a successful end state?
• User preferences & needs
   – Before they even start: coffee, pen, notebook, log sheets…

• Tasks & Subtasks:
   – Physical
   – Cognitive
   – Communication
• Conditions under which these tasks are done
• Results/outcomes of tasks
• Requirements to perform task:
   – Information
   – Communication with others
   – Equipment

Data Gathering Tools:

                Docs

• Documentation

   - Often contains description of how the tasks should be done (rather than how they are currently being done)
   - Standards
   - Manuals
   - Histories
   - Best Practices
• Domain Expert Description

  - Expert describes how process should work, how tasks should be done

  - “Knowledge-based” discovery

DGT: Interviews

• Interviews:
- Structured
  Efficient
  Require training
- Unstructured
  Inefficient
  No training
- Semi-structured
  Good balance
  Often appropriate

Semi-structured Interviews

• Predetermine data of interest
• Plan for effective question
types
     How do you perform task x?
     Why do you perform task x?
     Under what conditions do you perform task x?
     What do you do before you perform…?
     What information do you need to…?
     Who do you need to communicate with to…?
     What do you use to…?
     What happens after you…?
     What is the result or consequence of…?
     What is the result or consequence of NOT…? 

- See: Gordon & Gill, 1992; Graesser, Lang, & Elofson, 1987

DGT: Observation

• Observation
- In site, watch users do what they do
- Record with videotape
   To watch later, or again
- Take lots of notes, sketches
- May require coding the video later
- Focus on specific task-relevant behaviors in notes, but later
  convert to abstract subtasks

DGT: Questions, Think-aloud

• Questionnaires
   - Exploratory vs. confirmatory
   - Open-ended vs. categorical (exhaustive)
          What do you need to perform..? (list)
          Which of the following is most important to perform…? (select)
   - If you ask it, use it. If you won’t/can’t use it, don’t ask it.
• Think-aloud protocol
   - Person talks about what they are doing, while they are doing it (or just before or after)
   - Observer can ask probe questions
   - Why did you just do that?
• Note: Probe questions affect performance, as does thinking aloud.

DGT: Logging

• Automatic tracking
  - Keystroke/mouse click monitoring
  - Timers
  - Logs
  - Physical location/movement trackers
    Cell phones
    Aware Home

Representing Data:
       Outlines

• Lists, outlines, matrices
  - Use expanding/collapsing outline tool
  - Add detail progressively
  - Know in advance how much detail is enough
  - Can add linked outlines for specific subtasks
  - Good for sequential tasks
  - Does not support parallel tasks well
  - Does not support branching well

 

RD: Narratives

• Narratives
  – Describe tasks in sentences
  – Often expanded version of list or outline
  – More effective for communicating general idea of task
  – Not effective for details
  – Not effective for branching tasks
  – Not effective for parallel tasks

RD: Hierarchies

• Hierarchical Task Analysis (HTA)
  – Graphical notation & decomposition of tasks
  – Tasks as sets of actions
  – Tasks organized into plans
   Clusters of subtasks with a preferred order
   and prerequisite conditions
• Example Task Clusters
  – Fixed sequence
  – Optional tasks
  – Waiting events
  – Cycles
  – Time-sharing
  – Discretionary

RD: Networks

• Network / Entity-Relationship Diagrams
– Objects/people with links to related objects
  Stress relationship between objects and actions
– Links described functionally and in terms of strength
  Task: Develop design for final project
       objects - pens, paper, drawing tools, etc.
       actors - Mary, Bob, Sally
       composite objects - the “team”
– Often list attributes, actions of objects
         Object: pen simple
           Attribute:
              color: red
              writing: on/off
 

         Object: Mary actor
           Actions:
              M1: make a sketch
              M2: organize meeting

RD: Flow Charts

• Flow Chart of Task Steps
  – Combines Entity-relationship (network) with sequential flow, branching, parallel tasks.
  – Includes actions, decisions, logic, by all elements of the system
  – Abstracted
  – Mature, well-known, good tools