Requirements Gathering

1
Requirements Gathering

• CSCI 4800/6800
• Feb. 25, 2003

2
Goals

• What does the client want/need? (goal go from
  vague to specific)
• Representation of problems with current system
• Representation of requirements of new system

3
Techniques

• Interviewing/questionnaires
• Observation
• Document analysis
• Prototyping
• Checklists

4
Types of Requirements

• Functional - what system must do
• Data - structure of system or data neccessary
• Usability - acceptable levels of user performance
  satisfaction

5
Functional Requirements

• Both human system
• Dataflow diagrams
• Data dictionaries

6
Data Requirements

• Entity-Relationship diagrams

7
Usability requirements

• Learnability
• Throughput
• Flexibility
• Attitude

8
Functional Requirements

• What the system does
• What the user does
• Result of analysis collecting requirements is
  functional specification
• Dataflow diagrams
• Document or other representation
• Separate modules
• Organized hierarchically

9
Functional Requirements

• Databases - more persistent than dataflows
• An abstraction of the existing system, not a copy
  of the current physical process
• Basis for structured walkthrough - verbal
  description of the system based on the
  diagrammatic representation
• Other approaches
• Flow charts
• Scenarios - good for HCI

10
Data Requirements

• Attention on structure rather than processing
• Elicit using
• Observation
• Document analysis
• Interviewing
• etc.

11
Data Requirements

• Entity-Relationship (ER) diagrams used with
  formal descriptions of data elemtents, entities,
  relationships kept in a data dictionary, to
  describe structure context of the data in a
  system
• Entity - aggregation of a number of data elements
  (attributes of the entity)
• Entity type - classification according to shared
  attributes/attribute types
• Relationships - association between two or more
  entities

12
Usability requirements

• Easy to learn remember, Useful, Easy pleasant
  to use
• Learnability
• Time/effort to reach specified level
• Throughput
• Tasks accomplished by experienced users
• Speed of task execution
• Errors made
• Flexibility
• Extent to which system can accommodate changes
• Attitude

13
Determining usability requirements

• Task analysis (next section)
• determine cognitive and other characteristics
  required of users by system (search strategy,
  prereq knowledge, cognitive loading, etc.)
• User analysis
• determine scope of population who will use the
  system user modeling techiques may be applied
  here typically use checklists
• Environment analysis
• where system will operatre

14
Task Analysis

• determine cognitive and other characteristics
  required of users by system
• search strategy
• prereq knowledge
• cognitive loading
• etc.

15
User Analysis

• User analysis
• determine scope of population who will use the
  system
• user modeling techiques may be applied here
• typically use checklists

16
Environment Analysis

• where system will operate
• Physical aspects
• User support environment
• more

17
Usability requirements

• Expressed as usability metrics
• Completion time for specific tasks by specific
  set or type of user
• Number of errors per task
• Time spent using documentation

18
Usability metrics

• time to complete task
• of task complete
• complete per unit time
• successfailure ratio
• time spent on errors
• number of errors
• number of commands used
• frequency of doc/help use

• time spent in help/doc
• favorableunfavorable comments
• number of reps. of failed commands
• number of successful runs
• number of failure runs
• number of times interface misleads user
• .... more

19
Components of usability

• Learnability
• time and effort required to reach a specified
  level of use performance (ease of learning)
• Throughput
• tasks accomplished by experienced users
• speed of task execution, number and type of
  errors
• Flexibility
• extent to which system can accomodate change
• Attitude
• do users like it?

20
Usability study

• the act of gathering usability requirements
• involves
• task analysis
• user analysis
• environment analysis

21
Task analysis

• determine characteristics, particularly
  cognitive characteristics, required of users by
  system
• search strategy
• pre-req knowledge
• cognitive loading

22
User analysis

• determine scope of user population that will use
  system
• intellectual ability, cognitive processing
  ability, previous experience, physical
  capabilities
• typically a checklist description, rather than
  chart or diagram as in dataflow and E-R

23
Environment analysis

• physical environment
• temperature, humidity, lighting, available space
• social environment
• quiet/loud, busy/isolated, etc.
• user support environment
• helpful/hostile