WEBVTT

00:00:11.160 --> 00:00:14.160 align:center line:-2
Hello. My name is
Alejandro Rodríguez Ascaso.

00:00:14.320 --> 00:00:18.080 align:center line:-2
I am a lecturer at the Department
of Artificial Intelligence at UNED.

00:00:18.120 --> 00:00:20.560 align:center line:-2
This is the second video
in the series

00:00:20.600 --> 00:00:23.160 align:center line:-2
about the development
of user requirements,

00:00:23.200 --> 00:00:26.280 align:center line:-2
and it deals with accessibility
and usability needs.

00:00:27.520 --> 00:00:30.680 align:center line:-2
Specifically, today we are going
to work through an example,

00:00:30.720 --> 00:00:32.640 align:center line:-1
an example applied to a task.

00:00:33.080 --> 00:00:38.080 align:center line:-2
A task consisting of the use
of a voice recogniser to edit text,

00:00:38.280 --> 00:00:41.480 align:center line:-2
to precisely improve usability
and accessibility.

00:00:44.520 --> 00:00:46.080 align:center line:-1
The profile we are going to address

00:00:46.120 --> 00:00:48.880 align:center line:-2
 is that of a user
with little manual dexterity,

00:00:49.040 --> 00:00:51.520 align:center line:-2
and the use case
is that of editing text;

00:00:51.560 --> 00:00:53.040 align:center line:-1
and within that use case,

00:00:53.080 --> 00:00:56.960 align:center line:-2
one step in the basic course of action,
namely dictating a paragraph.

00:00:58.360 --> 00:01:00.320 align:center line:-1
Let us recall the interaction sequence.

00:01:00.480 --> 00:01:05.080 align:center line:-2
In it, the first thing is to perceive
the existence and position of controls.

00:01:05.200 --> 00:01:08.280 align:center line:-2
Then, to perceive the status
of those system controls.

00:01:08.720 --> 00:01:11.400 align:center line:-2
Then, to activate
the system controls.

00:01:11.840 --> 00:01:15.680 align:center line:-2
Then, to perceive the feedback
associated with that activation.

00:01:15.840 --> 00:01:17.600 align:center line:-1
And the cycle begins again,

00:01:17.640 --> 00:01:20.600 align:center line:-2
without forgetting that the user
must be able to understand

00:01:20.640 --> 00:01:22.640 align:center line:-2
both the output
produced by the system

00:01:22.680 --> 00:01:25.840 align:center line:-2
 and, in general, how to use
that product or system.

00:01:32.160 --> 00:01:36.120 align:center line:-2
Let’s begin with perceiving
the existence and position of controls.

00:01:36.600 --> 00:01:38.880 align:center line:-1
We are in a voice-recognition system.

00:01:39.320 --> 00:01:41.480 align:center line:-1
Therefore, it does not make sense

00:01:41.520 --> 00:01:44.000 align:center line:-2
to expect to perceive
the controls visually

00:01:46.680 --> 00:01:51.040 align:center line:-2
or learn from the system
how those controls are arranged.

00:01:51.080 --> 00:01:54.400 align:center line:-2
Rather, the user needs to know
which voice commands

00:01:54.440 --> 00:01:56.720 align:center line:-1
can be spoken at each moment.

00:02:00.000 --> 00:02:03.320 align:center line:-2
If we look up
the Dragon user manual online,

00:02:03.360 --> 00:02:06.760 align:center line:-2
 we can see that there is
a command called "What can I say?"

00:02:06.800 --> 00:02:08.880 align:center line:-1
which opens Dragon's sidebar,

00:02:08.920 --> 00:02:12.760 align:center line:-2
where the commands the user
may use at that moment are shown.

00:02:14.440 --> 00:02:16.800 align:center line:-1
We can now start to fill in our table.

00:02:17.200 --> 00:02:20.360 align:center line:-2
First, we indicate the need
we are identifying,

00:02:20.560 --> 00:02:25.080 align:center line:-2
in this case,
accessibility principle ACC-3-8:

00:02:26.000 --> 00:02:28.360 align:center line:-2
information that describes
the arrangement

00:02:28.400 --> 00:02:31.600 align:center line:-2
of the parts of the system
that enable its operation.

00:02:32.280 --> 00:02:35.080 align:center line:-2
According to the user manual,
this can be accessed

00:02:35.120 --> 00:02:37.280 align:center line:-1
through the "What can I say?" command, 

00:02:37.320 --> 00:02:39.880 align:center line:-2
which allows the list
of voice commands

00:02:39.920 --> 00:02:42.080 align:center line:-2
that the user
can enunciate at that moment

00:02:42.240 --> 00:02:43.600 align:center line:-1
to be shown visually.

00:02:43.880 --> 00:02:47.040 align:center line:-2
This same principle
also satisfies recognition

00:02:47.080 --> 00:02:48.520 align:center line:-1
rather than recall

00:02:48.560 --> 00:02:51.200 align:center line:-2
according
to Nielsen's usability heuristics,

00:02:51.240 --> 00:02:53.840 align:center line:-2
 so we also write that down
in the table.

00:02:54.840 --> 00:02:58.600 align:center line:-2
Next, after perceiving the existence
and position of controls,

00:02:58.720 --> 00:03:01.600 align:center line:-2
comes perceiving
the status of those controls.

00:03:05.640 --> 00:03:08.960 align:center line:-2
When Claudia
is dictating text to her computer,

00:03:09.000 --> 00:03:12.360 align:center line:-2
we see that at all times
there is an icon, a symbol,

00:03:12.480 --> 00:03:16.000 align:center line:-2
indicating to Claudia
that the voice-recognition system

00:03:16.040 --> 00:03:19.520 align:center line:-2
is ready to listen
and transcribe her voice into text.

00:03:25.280 --> 00:03:29.000 align:center line:-2
Indeed,
accessibility principle ACC-4-5

00:03:29.040 --> 00:03:31.760 align:center line:-2
states that there should be
a visual indicator

00:03:31.800 --> 00:03:34.520 align:center line:-2
that is visible to someone
with visual impairment.

00:03:35.520 --> 00:03:37.960 align:center line:-2
 Is it visible to someone
with visual impairment?

00:03:38.080 --> 00:03:39.280 align:center line:-1
That is not certain.

00:03:39.320 --> 00:03:43.360 align:center line:-2
Our recommendation would be
to increase its size so that any user,

00:03:43.480 --> 00:03:46.280 align:center line:-2
 regardless of their visual ability,
can perceive it. 

00:03:46.560 --> 00:03:49.080 align:center line:-2
For example, if I take off
my glasses right now,

00:03:49.120 --> 00:03:52.360 align:center line:-2
I am unable to see
what is currently on the screen.

00:03:52.840 --> 00:03:56.120 align:center line:-2
So we must ensure
that visually presented information

00:03:56.160 --> 00:03:58.080 align:center line:-1
is adequate for any user,

00:03:58.120 --> 00:04:01.080 align:center line:-2
or at least for the greatest number
of users possible.

00:04:03.280 --> 00:04:05.200 align:center line:-1
For this reason, in this case,

00:04:05.240 --> 00:04:08.720 align:center line:-2
the characteristic we want
when preparing the user requirements

00:04:08.760 --> 00:04:11.400 align:center line:-2
is that we would like
to increase the size.

00:04:11.440 --> 00:04:12.760 align:center line:-1
What are the implications?

00:04:13.680 --> 00:04:16.920 align:center line:-2
Well, increasing
the size of the indicator

00:04:16.960 --> 00:04:20.000 align:center line:-2
might make it interfere
with the rest of the visual information

00:04:20.040 --> 00:04:21.960 align:center line:-1
displayed by the word processor. 

00:04:22.320 --> 00:04:27.200 align:center line:-2
At this stage of user-centred design,
we simply note this implication,

00:04:27.240 --> 00:04:30.120 align:center line:-2
and later, the engineering team
working to improve

00:04:30.160 --> 00:04:33.640 align:center line:-2
the accessibility and usability
of this system will examine it.

00:04:40.440 --> 00:04:44.600 align:center line:-2
Another principle applicable in this case
is that controls and indicators

00:04:44.640 --> 00:04:48.480 align:center line:-2
must be perceptible without needing
to perceive colour differences.

00:04:48.520 --> 00:04:51.840 align:center line:-2
In this case, colour
does not encode any information.

00:04:53.880 --> 00:04:56.840 align:center line:-2
We continue with needs
related to visual perception,

00:04:57.120 --> 00:05:01.040 align:center line:-2
in this case, adequate brightness
in the information presented.

00:05:03.120 --> 00:05:06.920 align:center line:-2
The desired characteristic
is that sufficient brightness exists,

00:05:07.200 --> 00:05:09.800 align:center line:-2
and usually, the brightness
of a computer screen

00:05:09.840 --> 00:05:11.640 align:center line:-1
is controlled by the user.

00:05:11.920 --> 00:05:15.560 align:center line:-2
It will be necessary to check
that the option to control brightness

00:05:15.600 --> 00:05:18.440 align:center line:-2
in the operating system
is accessible to the user.

00:05:18.640 --> 00:05:21.720 align:center line:-2
Although this is not the subject
of the design of our system,

00:05:21.800 --> 00:05:24.400 align:center line:-2
but rather the design
of the operating system

00:05:24.440 --> 00:05:26.200 align:center line:-1
or the computer itself.

00:05:33.720 --> 00:05:38.960 align:center line:-2
Next comes another type of need related
to the perception of visual information.

00:05:39.560 --> 00:05:42.440 align:center line:-2
In this case, information
within the visual field

00:05:42.480 --> 00:05:46.440 align:center line:-2
of people who are of short stature
or seated in a wheelchair.

00:05:47.200 --> 00:05:48.720 align:center line:-1
The desired characteristic

00:05:48.760 --> 00:05:51.360 align:center line:-2
 is that the user is
in the appropriate position.

00:05:51.440 --> 00:05:54.600 align:center line:-2
We note, as an explanation,
that this must be ensured,

00:05:54.960 --> 00:05:57.160 align:center line:-2
although again,
this is not a characteristic

00:05:57.200 --> 00:05:59.400 align:center line:-2
that affects the design
or improvement

00:05:59.440 --> 00:06:01.440 align:center line:-1
of our voice-recognition system.

00:06:07.360 --> 00:06:11.080 align:center line:-2
Similarly, in the case
of avoiding reflections or glare,

00:06:11.520 --> 00:06:14.440 align:center line:-2
when perceiving visual information
in the foreground

00:06:14.480 --> 00:06:16.200 align:center line:-1
while there is a background,

00:06:16.240 --> 00:06:20.000 align:center line:-2
it is the task of those who design
Dragon NaturallySpeaking.

00:06:23.480 --> 00:06:28.040 align:center line:-2
What we can do at this stage is state
that contrast must be sufficient, 

00:06:28.080 --> 00:06:30.960 align:center line:-2
and note as an implication
that there are tools

00:06:31.000 --> 00:06:33.760 align:center line:-2
to check
whether the contrast is sufficient.

00:06:35.320 --> 00:06:40.280 align:center line:-2
For example, tools
such as WCAG contrast checkers

00:06:41.360 --> 00:06:44.480 align:center line:-2
can verify whether
an application's contrast

00:06:44.520 --> 00:06:48.320 align:center line:-2
complies with
Web Content Accessibility Guidelines.

00:06:53.560 --> 00:06:55.680 align:center line:-1
Let's continue with our sequence.

00:06:55.800 --> 00:07:00.280 align:center line:-2
After perceiving the status of controls
comes the activation of controls.

00:07:02.840 --> 00:07:06.680 align:center line:-2
In this case, the characteristic
we want is that the user,

00:07:06.800 --> 00:07:10.800 align:center line:-2
who has little manual dexterity
and not much movement precision,

00:07:11.040 --> 00:07:12.600 align:center line:-1
can use the system.

00:07:14.720 --> 00:07:18.160 align:center line:-2
Well, that is precisely
why we have chosen this system.

00:07:18.200 --> 00:07:19.880 align:center line:-1
Claudia can use this product

00:07:19.920 --> 00:07:23.000 align:center line:-2
because she can dictate
the paragraph with her voice.

00:07:24.640 --> 00:07:28.600 align:center line:-2
As an explanation, I note
that it will be necessary to ensure

00:07:28.640 --> 00:07:32.800 align:center line:-2
that the user is in an environment
without sound interference.

00:07:32.880 --> 00:07:36.200 align:center line:-2
Again, this is not something
we need to address in the design

00:07:36.240 --> 00:07:39.000 align:center line:-2
of a new version
of our voice-recognition system,

00:07:39.040 --> 00:07:41.200 align:center line:-1
but there is no problem in noting it.

00:07:45.440 --> 00:07:47.360 align:center line:-1
After activating the control,

00:07:47.400 --> 00:07:50.480 align:center line:-2
we want
the corresponding perception of feedback.

00:07:51.600 --> 00:07:55.200 align:center line:-2
In this case, feedback
is provided by Microsoft Word,

00:07:55.320 --> 00:07:59.440 align:center line:-2
writing in the document the sentence
that Claudia has just pronounced,

00:08:00.000 --> 00:08:02.160 align:center line:-1
and that acts as feedback.

00:08:08.320 --> 00:08:10.760 align:center line:-2
What we need is
for the visual feedback

00:08:10.800 --> 00:08:13.960 align:center line:-2
to be evident for people
with visual impairment.

00:08:14.360 --> 00:08:15.680 align:center line:-1
We write this down.

00:08:16.400 --> 00:08:18.320 align:center line:-2
The feedback,
the text of the editor,

00:08:18.360 --> 00:08:21.720 align:center line:-2
must be checked to ensure
that its visual characteristics

00:08:21.760 --> 00:08:23.880 align:center line:-1
can be controlled by the user.

00:08:23.920 --> 00:08:27.800 align:center line:-2
Again, this is not a specific task
of the voice recogniser

00:08:27.840 --> 00:08:31.240 align:center line:-2
but of the application
that the recogniser is controlling.

00:08:46.800 --> 00:08:49.400 align:center line:-1
Then, after perceiving the feedback,

00:08:49.920 --> 00:08:53.720 align:center line:-2
the perception of the existence
and position of controls begins again.

00:08:59.280 --> 00:09:03.240 align:center line:-2
What does the user do?
She says, "Start new line."

00:09:03.560 --> 00:09:08.080 align:center line:-2
Here we are in another use case,
we've called it "Start new line",

00:09:09.160 --> 00:09:12.720 align:center line:-2
and in that use case,
practically all the user requirements

00:09:12.760 --> 00:09:15.680 align:center line:-2
are the same as the ones
we have already created.

00:09:17.280 --> 00:09:20.840 align:center line:-2
We see one small difference:
in the feedback of the operation,

00:09:20.880 --> 00:09:23.400 align:center line:-2
here the previous symbol
no longer appears,

00:09:23.440 --> 00:09:26.440 align:center line:-2
but instead the phrase
associated with the command

00:09:26.480 --> 00:09:30.600 align:center line:-2
that Claudia has just spoken appears:
"Start new line."

00:09:31.480 --> 00:09:34.680 align:center line:-2
That probably leads us
to introduce some new principle

00:09:35.000 --> 00:09:37.760 align:center line:-2
relating to the need
to perceive feedback

00:09:37.920 --> 00:09:40.720 align:center line:-2
associated with text
in this specific case.

00:09:46.880 --> 00:09:49.400 align:center line:-2
But let us not forget
where we came from.

00:09:49.560 --> 00:09:52.040 align:center line:-2
We have now completed
our full sequence,

00:09:52.080 --> 00:09:54.480 align:center line:-2
but we are still missing
something very important:

00:09:54.520 --> 00:09:55.720 align:center line:-1
comprehension,

00:09:55.760 --> 00:09:58.120 align:center line:-2
understanding both
the output of the system

00:09:58.160 --> 00:10:00.000 align:center line:-1
and the functioning of the system.

00:10:00.040 --> 00:10:02.800 align:center line:-2
Let us analyse all the needs
step by step.

00:10:04.400 --> 00:10:08.960 align:center line:-2
On the one hand, the usability
principle help section and documentation

00:10:09.000 --> 00:10:12.760 align:center line:-2
are satisfied because Claudia
tells us at the beginning of the video

00:10:12.800 --> 00:10:16.400 align:center line:-2
that there is user help
that is useful and should be used.

00:10:16.840 --> 00:10:20.960 align:center line:-2
On the other hand,
the accessibility principle ACC-3-8

00:10:21.000 --> 00:10:23.320 align:center line:-1
states that the information describes

00:10:23.360 --> 00:10:25.440 align:center line:-2
the arrangement
of the parts of the system

00:10:25.480 --> 00:10:27.080 align:center line:-1
that allow its operation.

00:10:27.440 --> 00:10:28.600 align:center line:-1
We saw that earlier.

00:10:28.640 --> 00:10:30.320 align:center line:-1
There is also contextual help:

00:10:30.360 --> 00:10:33.280 align:center line:-2
 when the user says
the command "What can I say?",

00:10:33.480 --> 00:10:36.480 align:center line:-2
the list of voice commands
that can be spoken at that moment

00:10:36.520 --> 00:10:37.760 align:center line:-1
is shown visually.

00:10:38.400 --> 00:10:40.960 align:center line:-2
We also saw earlier
that this satisfied

00:10:41.000 --> 00:10:44.440 align:center line:-2
Nielsen's heuristic principle
of recognition rather than recall.

00:10:46.480 --> 00:10:49.680 align:center line:-2
Let's continue with other principles
related to comprehension,

00:10:49.720 --> 00:10:53.960 align:center line:-2
for example, that text, symbols, etc.,
should be easy to understand.

00:10:54.400 --> 00:10:57.760 align:center line:-2
In this case,
the command texts seem appropriate.

00:10:58.200 --> 00:11:02.520 align:center line:-2
It seems logical that when the user
says "Start new line,"

00:11:06.760 --> 00:11:10.600 align:center line:-2
it is a command that is quite easy
to remember and understand,

00:11:10.880 --> 00:11:13.760 align:center line:-2
precisely because the text
corresponds directly

00:11:13.800 --> 00:11:15.080 align:center line:-1
to what we want to do:

00:11:17.160 --> 00:11:18.680 align:center line:-1
start a new line.

00:11:21.960 --> 00:11:24.680 align:center line:-2
We do not like
the symbol used to indicate

00:11:24.720 --> 00:11:27.040 align:center line:-2
that voice recognition
is ready to listen.

00:11:27.360 --> 00:11:30.440 align:center line:-2
It is a symbol that does not
necessarily need to be good.

00:11:30.480 --> 00:11:32.880 align:center line:-1
It must be easily recognised by the user. 

00:11:33.400 --> 00:11:35.880 align:center line:-2
We might therefore want
either to use a symbol

00:11:35.920 --> 00:11:37.800 align:center line:-1
that is internationally recognised

00:11:37.840 --> 00:11:40.840 align:center line:-2
as meaning
that a voice-recognition system is active,

00:11:41.400 --> 00:11:43.160 align:center line:-1
if such a symbol exists,

00:11:43.240 --> 00:11:45.480 align:center line:-2
we do not know,
we could look it up,

00:11:48.920 --> 00:11:52.840 align:center line:-2
or to offer users the possibility
of customising these contents.

00:11:53.800 --> 00:11:57.120 align:center line:-2
The implication is
that customising these contents

00:11:57.160 --> 00:12:00.240 align:center line:-2
might increase the complexity
of the user interface,

00:12:00.520 --> 00:12:01.680 align:center line:-1
because it adds

00:12:01.720 --> 00:12:04.280 align:center line:-2
another configuration question
for the user.

00:12:05.040 --> 00:12:07.400 align:center line:-2
This content
that we have entered in the table

00:12:07.440 --> 00:12:09.480 align:center line:-1
also applies to other principles,

00:12:09.520 --> 00:12:14.040 align:center line:-2
for example, that products or services
use standard conventions, words,

00:12:14.080 --> 00:12:16.360 align:center line:-2
or symbols
from the user's own culture;

00:12:16.400 --> 00:12:20.160 align:center line:-2
that text materials are written
as clearly and simply as possible,

00:12:20.200 --> 00:12:24.600 align:center line:-2
and the usability principle of match
between the system and the real world.

00:12:26.080 --> 00:12:27.680 align:center line:-1
There is also another principle

00:12:27.720 --> 00:12:30.040 align:center line:-2
relating to the comprehension
of the information

00:12:30.080 --> 00:12:31.400 align:center line:-1
displayed by the system:

00:12:31.440 --> 00:12:35.360 align:center line:-2
that text, illustration, and diagrams
have a narrated format.

00:12:35.640 --> 00:12:38.760 align:center line:-2
 This would improve comprehension
for some types of users.

00:12:38.800 --> 00:12:40.720 align:center line:-2
But as an implication,
we must note

00:12:40.760 --> 00:12:42.880 align:center line:-1
that generating audio information

00:12:42.920 --> 00:12:45.640 align:center line:-2
might interfere
with voice-dictation interaction,

00:12:45.680 --> 00:12:49.520 align:center line:-2
because the voice recogniser
might pick up that audio information,

00:12:49.560 --> 00:12:51.960 align:center line:-2
or even if the user
listens with headphones,

00:12:52.080 --> 00:12:54.720 align:center line:-2
they may have difficulty
generating voice commands

00:12:54.760 --> 00:12:56.480 align:center line:-1
while simultaneously listening

00:12:56.520 --> 00:12:59.400 align:center line:-2
to that audio information
produced by the system.

00:13:02.440 --> 00:13:05.680 align:center line:-2
To conclude, we can say
that we have created

00:13:05.720 --> 00:13:08.840 align:center line:-2
user requirements
for usability and accessibility,

00:13:08.880 --> 00:13:10.960 align:center line:-2
covering the following steps
of the sequence:

00:13:11.000 --> 00:13:13.240 align:center line:-2
perceive, activate,
and understand.

00:13:13.360 --> 00:13:15.080 align:center line:-1
But we are still missing some:

00:13:15.120 --> 00:13:17.560 align:center line:-2
equivalent safety
and privacy requirements,

00:13:17.600 --> 00:13:20.520 align:center line:-2
avoiding personal risks,
using the product efficiently,

00:13:20.560 --> 00:13:23.880 align:center line:-2
using assistive products,
and other cross-cutting issues.

00:13:23.920 --> 00:13:25.560 align:center line:-1
See you in the next video.