The test preparation and planning support module assists in the development of usability test plans and data collection forms. The module tailors and formats participant demography data forms, subject consent to participate forms, task checklists (supporting generations of test scenarios), laboratory setup checklists, and scenario development. The module outputs hard copies of forms and data that can be edited using a text editor, or used directly in test plans. The tool contains a outline of a generic usability test that can be modified to meet most usability test planning needs.
The data logging and analysis module helps in the collection of task performance data and time-on-task data. This module defines specific keystrokes as the initiators and terminators of discrete task performed. Using easily definable function keys, task performance data can be collected in real time and subsequently analyzed by the software. During data collection, test personnel observe subjects and enter single keystrokes that begin time and task logging for each task in the taxonomy. Each incidence of the initiation of a task can be annotated either in real time or after testing. The software automatically analyses the data and reports:
• time spend performing each task category in terms of total time on each task and percent of total time spent on each task (e.g., "8% of total time spent in navigating screens")
• task counts and frequencies (e.g., "HELP queried a total of six time, or once every four minutes")
• task annotations (if entered in real time)
• error counts
The interface evaluation module allows the usability engineer to search a library of over 1000 user interface guidelines and apply any subset of them to any computer interface under evaluation. Complex search and selection algorithms are provided, as well the ability to print hard copies of evaluation checklists or to conduct evaluations "on-line." The software also allows for in depth of reporting the results of a guideline based usability evaluation. This module can be used to evaluate an interface where no actual usability study (task performance observation) is being performed.
APPROPRIATE USES:
Usability test planning (scenario generation, selection of performance measures, development of test procedures), usability test conduct (data logging, data reduction and computation of descriptive statistics), report generation, application of databased UCI standards to interfaces under development or test.
EQUIPMENT REQUIRED:
Apple® Macintosh running System 6.0 or higher. Hypercard 2.1 or later. Minimum of 8 megabytes of RAM, 3 Megabytes of hard disk drive space.
INPUTS REQUIRED:
The software guides the test developer through the development of test plans, to conduct of test, to data evaluation and reporting. Major input are descriptions of the device/software to be tested and use characteristics.
PROCESSING TECHNIQUES:
Task-activity logging, computation of descriptive statistics, report generation (event tables), data base search.
OUTPUT:
Task-activity logs, descriptive statistics, reports, objective measures of human performance using the device tested, subjective (user) measures, identified interface design improvements, usability predictions, inputs to user documentation, customer support plans, training requirements.
USES OF OUTPUT:
Interface redesign, problem reports, documentation inputs, quantified inputs to tradeoff decisions.
STAGE OF DEVELOPMENT:
Complete.
VALIDATION:
Statistical/computational modules validated analytically. Database, data logging, and reporting modules validated analytically. Planning modules validated subjectively.
COMMENTS: The importance of the design for usability in software development is evident in that: (a) the human computer interface comprises from 47% to 60% of the total lines of code; (b) a graphical user interface accounts for at least 29% of the software development budget; and (c) 80% of costs associated with the software life cycle (design, development, implementation, and maintenance and operation) accrue during the post-release maintenance phase of the life cycle, and furthermore, 80% of this maintenance is attributable to unmet or unforeseen user requirements. Therefore, 64% of the life cycle costs associated with a software system is due to changes required to improve the interface between user and computer.