Web Posted on: November 20, 1998

A SURVEY OF WINDOWS SCREEN READER USERS: RESULTS AND RECOMMENDATIONS

Crista Earl and Jay Leventhal
American Foundation for the Blind
11 Penn Plaza Suite 300
New York, NY 10001
(212) 502-7605
techctr@afb.net

With skill in the use of Windows now essential for employment, AFB's technology program has updated its earlier survey of Windows-based screen reader users to find out how blind and visually impaired computer users are faring. The results of the survey and recommendations for screen reader selection and development derived from the results and from screen reader evaluations conducted in AFB's Product Evaluation lab are presented.

The purpose of the survey reported here was to gather information from the user's perspective. The survey asked blind or visually impaired people who use Windows with screen readers what tasks they perform in Windows and how comfortable they feel performing those tasks.

The Survey

Users of Windows-based screen readers were selected from the American Foundation for the Blind's Careers and Technology Information Bank (CTIB) and contacted by E-mail and telephone. Over 400 persons were contacted, more than 200 of whom completed surveys. The survey was conducted from August 22 to October 15, 1998. Follow-up questions were sent to specific groups of participants after October 15, 1998.

Survey Questions

Respondents were asked what hardware and software they use; whether they use a braille display or screen magnification along with synthetic speech; what Windows or DOS applications they use; what methods they used to learn Windows; why they began using Windows; and what their overall rating is for how comfortable they feel in using Windows. Respondents were also asked if they were able to perform successfully each of a list of tasks in the Windows environment.

Participants

The survey participants were drawn from among the 526 CTIB members who use Windows screen readers. Of those 526, 81% have a college degree. Forty-One percent have a graduate degree. Eighty-eight percent are currently employed. 64% have no useful vision.

There are 44 computer programmers; 31 Assistive Technology Trainers; 20 attorneys; 11 rehabilitation counselors; 6 secretaries and 4 receptionists.

Clearly, the group from which the survey participants were drawn are highly successful visually impaired users. This group might be expected to use Windows applications and Windows screen readers with a higher level of success than what would be found in a random sampling of visually impaired computer users.

Responses

Hardware and software The respondents used the following Windows-based synthetic speech programs: ASAW (Automatic Screen Access for Windows) from MicroTalk Software; JAWS (Job Access with Speech) for Windows from Henter-Joyce; ScreenPower for Windows from TeleSensory Corp.; WinVision from Arctic Technologies; Window Bridge from Syntha-Voice Computers; Window-Eyes from GW Micro; outSPOKEN for Windows from ALVA Access Group. More than 21% of the respondents reported using more than one Windows-based screen reader. Fifteen percent of respondents reported using a braille display in addition to synthetic speech. A small number of respondents used screen magnification along with synthetic speech.

Tasks Performed in Windows

Respondents were presented with a list of 22 Windows tasks and asked if each was something that they "do Easily," "do with difficulty," "cannot do," or had "never attempted." Since over 75% of the respondents submitted their responses via e-mail, it was not surprising that most respondents replied "Easily" to "reading and replying to e-mail messages." What was surprising was that over half of the respondents had difficulty or could not look up items in Windows help. Similarly, a disturbingly high percentage of respondents had difficulty filling out forms on the Internet, an essential skill for users for whom the Internet is the only means of access to otherwise printed materials.

Training

To compound the challenge that accompanies learning any new system, a large number of respondents had never had formal training. In 1996, 36% of the respondents had had some form of formal training. Among the respondents to the 1998 survey that percentage is considerably higher, but still disturbingly low. Further, not all respondents gave high marks to the formal training they did receive. "A general windows training class" was rated by the group far lower than "Books about windows." A task as daunting and visual as the Windows environment is much easier to master through formalized instruction than by individual trial and error.

What it all means

The results of this survey make clear that many blind or visually impaired employees believe that they need to use Windows to keep their jobs. A large number of the respondents said that they had to start using Windows at work or that they knew they would soon have to make the change from DOS.

At the same time, many participants reported that they still used DOS for some applications or for file management. Recommendations the authors were surprised by the frustration widely expressed by participants describing the performance of certain tasks, especially "Looking up items in Windows help," "Running a spell checker," and "Installing Windows applications ." At the same time, much of the frustration expressed in the 1996 survey is completely or nearly gone. Few respondents mentioned problems that could be interpreted as errors in their screen reader's off-screen model or other complete screen-reader failures. Only a few mentioned unlabeled graphics as problems. These observations lead the authors to conclude that the addition or enhancement of specific screen reader features and enhanced training would greatly smooth the blind Windows user's way.

Seven Windows-based screen readers have been evaluated in AFB's Product Evaluation lab within the past year and the best features of several of them will be demonstrated and discussed. These features have been selected for their ability to give access to programs, web sites, or situations in Windows which might otherwise not be useable by blind users. Ease of use and generalizability were considerations.