Before the
Federal Communications Commission
Washington, DC 20554

In the Matter of  Telecommunications Relay Services and Speech-to-Speech Services for
Individuals with Hearing and Speech Disabilities

 CC Docket No.98-67


I would like to make a few comments on TRS quality assurance parameters.
Quality Assurance Parameters
I am a deaf consumer and also a research scientist at the Univ. of Texas.
I use Relay Texas on a daily basis for professional and personal business.
I feel that the FCC should consider expanding the Quality Assurance
Parameters that are used to evaluate the performance of a TRS. At present
there is only one: the average speed of answer. I believe that it is not
possible to assure a quality TRS system with that single parameter, which
only gives a very vague indication of consumer satisfaction and TRS
performance. The number of parameters can easily be expanded and they can
easily be measured in an automated manner. This step will make great
strides in assuring uniform service across TRS providers.
A necessary part of this quality assurance should entail actual test calls
made by an independent quality assurance organization. These test calls
are the only way to insure that the TRS providers are actually providing
quality service and would provide the FCC with a quanititative measure of
performance. Relying on hearsay and anecdotes is not sufficient.
I do not have the background to make a comprehensive list of suggested
quality assurance parameters but I would like to mention one informal one
that we tested in Texas that led to some positive changes. Several deaf
consumers had complained for some time that when they requested VCO during
a call that it took too long for the agent to switch and respond to the
consumer, 'Voice now, please', which is what the Sprint protocol calls for.
In addition, some agents responded with different comments (some quite
lengthy), others just said 'GA', and others NEVER responded. I suggested
to Sprint that they do a set of 20 test calls measuring the time between
the request for VCO and the response. I also made 20 test calls to measure
the same parameter. To make a long story short, my results and Sprint's
results agreed: the response time varied tremendously from 1 sec to as
long as 20 sec , the average was far higher than what Sprint expected, and
the response text varied from one agent to the next. As a result of this,
Sprint implemented several training changes and redid the test. The retest
showed substantial improvement and Sprint has agreed to redo the test
periodically to insure this performance. Sprint also agreed to set up a
separate 800 number for incoming VCO calls to address this problem (and
other related VCO problems). The point of the story is that these changes
were not made until we had QUANTITATIVE EVIDENCE that the problem existed.
VCO consumers had complained for several years before this with anecdotes
of poor performance but the impetus for change came from this test.
There are a number of other parameters that can be measured that can lead
to this type of improvement.