A new resource: evaluate the useability of the software you’re developing

By: Carol Adams

If you develop hardware, software, or information, then you already know how important usability is to your product’s success. Whether your customers are experts or novices, dedicated users or occasional ones, good usability will improve their satisfaction and productivity… and make you more money.

Good usability means lower support costs and fewer post-ship changes. You’ll also benefit from positive trade press reviews, glowing user testimonials, and emphasizing ease of use in your advertising campaign. But how do you know when you have it?

By using the appropriate measurement techniques, Human Factors Engineers and other usability professionals can measure ease of use quite reliably. These measurements can be made at virtually any stage of product development, from conceptual design through end-of-life. Early measurements rely on low-fidelity prototypes. Testing later-on uses higher fidelity prototypes or any level of the product on hand.

The first of three basic concepts underlying usability measurement is that the measurements are meaningful only in the context of a given type of user performing a given task in a given product environment. In other words, regardless of how we define usability and what we decide to measure, our observations will have practical significance only for a specified combination of user, task, and product.

The second basic concept is that usability can be operationally defined in terms of user performance and satisfaction: we specifically assume that a particular combination of performance and satisfaction scores define "usability" sufficiently for our purposes.

The third concept is that both user performance and satisfaction can be measured in the laboratory or in the field, in the context of a given user type, task, and product. For example, user performance is frequently characterized in terms of time on task, errors, and so on, while satisfaction is best characterized in terms of scores on psychometrically validated questionnaires. Thus performance can be measured objectively by time-logging specific behavioral events as a user performs a task with a product; examples of these events include beginning and ending the task, beginning and ending error states, and so on. And during or after performance testing, satisfaction can be measured by asking the user for structured self-reports on satisfaction with main product dimensions (e.g., hardware, software, information).

A related and highly useful concept is that the usability of two or more products can be compared by holding user and task characteristics constant across products and taking performance and satisfaction measurements in a consistent manner. In fact, when any two of the three main factors in the usability equation–user, product, and task–are held constant and the third is varied, the effects on usability of the factor that is varied can be measured and assessed.

Human factors measurement techniques can also be applied to provide quantitative data on the usability of competitive products. Developers can use such data to determine whether their product’s usability is something fixed or featured, to generate measurable usability requirements, or to establish a baseline for measuring changes in usability as a function of changes in design.

Carol A. Adams is a Certified Professional Ergonomist and trained Human Factors Engineer with 18 years’ experience in computer usability. Previously an employee of The MITRE Corporation, IBM, and HaL Computer Systems, she is currently President of ergosoft, inc., a human factors consulting group headquartered in Austin with clients in Central Texas and Silicon Valley.

With decades of experience in computer usability, the ergosoft team can help you deliver products that are simple to install, intuitive to learn, easy to use, and effortless to maintain. Offering a complete line of professional usability services to hardware, software, and information developers, ergosoft can satisfy any usability outsourcing need.