Cognitive Ergonomics

Dr. Manish Kumar Asthana

Department of Humanities and Social Sciences
Indian Institute of Technology Roorkee, Uttarakhand
Lecture-X

Human Computer Interaction-II

Hello everyone. I welcome you all today in the lecture series of cognitive ergonomics. Today
in this lecture series, we will start from where we left off in the previous lecture. In the previous
lecture, we started to talk about the human-computer interaction. We discussed about the
human. We discussed about the interface.

We discussed about how different paradigms and interfaces are coming all together to assist
us in the field of human-computer interaction. We also understood in previous lecture that
personal computing, decentralized computing, how it helped exponentially in the discipline of
human-computer interaction. Now what we see here is the same design which I showed you
in the previous lecture is that the humans are there when humans are being involved here, then
the several cognitive components are involved here and these cognitive components are then
through their sensory system, physical system are interacting with the hardware, software of a
system. Now the interaction among them resulting into also in an helping us in understanding
the artificial intelligence, how we can come up in an artificial intelligence in a neural network,
building expert system, robotics, machine learning and natural language processing.

This has to be the human centered approach and now based on the interaction, the way human
and machines are interacting, the way humans and computers are interacting, this is also an
interaction design and interaction-oriented design and here it is the human oriented design.
There are different places where humans are interacting with computers. All these different
contexts has its own challenges and provides an aid to the user. For instance, if you see this air
traffic controller, the air traffic controller on the ground is coordinating and collaborating with
the aviators who are flying airplane in the sky. Lot of simulation, lot of understanding, lot of
complex cognitive task are happening here.

The important thing here is that with what easiness, with how much easiness these operators
are interacting with these systems and with what easiness these operators are capable enough
to solve the problem sitting on the ground of the aviator. Such challenges are not, do not exist
only in this discipline or in this context, but there are many more contexts where the operator,
operators task is very challenging and complex in nature. Certain skills and capabilities which
is required by the operator to operate the complicated and sophisticated system requires some
level of support and easiness. This help and support what we are talking about can be provided
with the easiness between the interaction between the user and the system. Now, the operator
when he is dealing with the sophisticated system of like fMRI, then not only he has to
understand how the system works, but also what is the output and what is the interpretation of
the output as well.

Such understanding, such interface interaction is highly demanding. Such interface

interaction is very challenging. And because of this, in previous lecture we also discussed
about the schema map, cognitive map, mental models and how they are contributing and
helping in the human cognition. How the cognition of the person, the operator, the analyst is
going to influence the performance efficiency and satisfaction at workplace setting. If you see
the naval officer, not only he has to navigate himself in the tide or under no tide condition, but
he has to coordinate with the other on-board employees.

Such coordination requires high level of cognitive processes. Such are very common
examples towards us which is giving an idea all together the importance and relevance of
cognition in ergonomics. When we alone talk about the human factor itself, then we understood
very soon that human computer interaction, human's role are very much important in this
interface. And when the human computer interaction came into existence from 1970 to 1980,
since then last 4 decades, things have changed drastically. There are several design principles
which people have talked about.

And in this lecture series, we are going to address such small components of interface, small
components of design, how they are playing a crucial role and helping in the efficiency and
performance of the user. Now in the software interface design, as we have been discussing
about, according to Schneidermann, the well-designed software interface can impact, a well-
designed interface can impact the learning rate at which person is learning. So, the interaction
which we are having today online, if this interface is not robust, if this interface is not so clear,
it may take more time to learn the same content related to a condition where the information is
presented in a very simpler form. We have to keep it extremely simple. Then only the
interaction between human and machine, human and computer can be enhanced and promoted.

When this interaction is easy and simple, then the learning time taken by a person is less.
Similarly, when the learning time is less, the performance is speed should be better. They
should be performing well. If they have learned well in less span of time because the task is
easy, the performance should be better. How can we measure the performance, accuracy and
reaction time? Now if the interface is easy, if the interface is effective, if interface is helpful
and providing aid to the user, then the rate of the error will be minimal.
 Lot of human errors have occurred because the interface is complicated for the user to
understand. In previous lecture, we have been discussing about the interface and we also took,
understood few examples of human error disasters such as Chernobyl, three mile radius, Bhopal
gas tragedy, etc. and etc. Here we are talking about this software interface. But the same
interface when a human is interacting with the machine in a big mechanical industry, and also
an interface is there.

It may not be a software interface, but it will be a physical interface which involves visual
ergonomics or visual communication. And if any visual representation, visual communication
is not up to the mark or do not have important characteristics, necessary characteristics, then
the chances of error will be high. Person will make an error. If the learning time is less, if the
performance speed is better, error rate is 0, then the user satisfaction will be high. So, error
rate is inversely proportional to user satisfaction.

Error rate is inversely proportional to learning time, lesser the learning time. Easy the learning
is minimal will be the error rate. Now performance speed if you talk about, performance speed
is better, error rate will be inversely proportional. So, software interface design if we are talking
about, there is a monetary saving, there is a consumer product, money is involved, product is
involved, but there is the possibility or probability of success and failure. The sophisticated
systems are there, such as we saw earlier in previous example, the air traffic control system
where the aviator in the space is coordinating with the control of the air traffic controller on the
ground, power utilities, naval navigation system, highly sophisticated, highly advanced
systems.

These systems provide a software interface design. It is very important for us to understand
that the softwares, the modern softwares are capable, are capable and also are having lot many
features, lot many life saving features which people are using and these features are providing
an aid to help the driver who is not very good, the intermittent driver to be an expert driver.
Also, there are extreme weather conditions when an aviator is flying an airplane. These
softwares can help in the critical scenarios. One thing we have to understand is that these
software in the modern world is of great importance and relevance.

People have been using these softwares last four decades to get an aid and assistance. Now
when we talk about a software design cycle, how these software is being designed, then we
have to understand the understanding behind the root cause of such a design, why such a design
is there, what kind of design we can come up with and then we have to evaluate also. We have
to understand why do we need that specific software. So, if we see the three levels, three cyclic
levels involved in the software design cycle, then first step is the understand and here what we
are able to see error criticality. Error is there, some errors are complicated and difficult to
understand, user experience and characteristics are there.
 We have to understand what my user has experienced, what kind of characteristics,
characteristic features my user is having. If he or she is capable to use such a software or such
a design, what experiences they have come up with. And finally what we have to also
understand is the organization and the culture where we are going to apply this software. The
second part is the computational and mental model. So, people are using several types of
computational model, machine learning to come close to the mental model.

When the designer model is close to the mental model of the user, then the success of the
design is very high. Moreover, we also need to come up with a theoretical framework why
such model is working, but why the other models will not work. We need a robust theoretical
framework so that we can understand the underlying root causes for such alteration and
changes. We can understand what is the need of this software design cycle. We can understand
what design cycle contribute.

We can understand what challenges design cycle is facing. We can understand how can we
make design cycle more robust. And finally, evaluation. Evaluation will be two types, based
on two types. One is the heuristic evaluation, the way different steps, different stages are
involved, one has to evaluate that.

Second, usability test and metrics. How much it is being usable? Is it effective to use? Does
it have any challenges to use? Now, we can combine these three parameters together. And you
can see it is cyclic in nature, which means once you understand the criticality of the error, why
the error has occurred, what is the underlying cause of the error? You use the patterns, you use
the goals, different goals, different functions, different tasks will be understood, will be
analyzed, will be investigated, user characteristics and expertise in which environment it is
being applied, organization and culture. Now, once we have these understanding, we move to
next level of design, where theoretical framework is required, computational model is required,
mental model and metaphors are required. Mental model, because we have a preconceived
notion or we expect to have a design which is similar to our mental model, consistent with our
mental model.

Metaphors can provide an easiness with which the information can be encoded easily.
Principles and guidelines, dialogue styles and standards. What kind of dialogue style is
required? What kind of principle is applied here? That those things we can. Later on, heuristic
evaluation, usability test and matrices, test and evaluation. Now, the coordination between
these two variables, we have to understand that people have tried to use all these parameters,
all these parameters to come to understand the user system.

Now, this picture some of you might be familiar with, some of you may not be. This is a word
document picture from Microsoft 1998. Now, 1998 Microsoft Word was having so many
features, more than half page is having the features. This type of featuring we call it as creepy
featuring, creeping features. What do we mean with that? It is creepy because all the
information is present there and now it is complicated for us to understand which feature is
important and which feature is not important.

This user system characteristics, if we talk about, we saw minutes ago about the user system,
creepism is there. So, the designer tried to reduce these content and today if you open the word,
you will find a one line. With very limited features, most of the features have gone into drop
down menu. Now, when we talk about user system and characteristics, functionality and use
of the use, the frequency of task performance. What is the frequency of task performance using
particular software? If you are using it frequently, then you may be an expert in this discipline.

Some features are mandatory, some features are discretionary use. So, this user system has
to propose which system is mandatory and which system is discriminatory use. Third,
knowledge level of the user. What is the knowledge level of the user? All these four
components are going to help us in understanding the user system and characteristics. Now, as
you can see on your left-hand side, this first page 2000 is been divided into easy, expert, normal
and hardcore.

Now, with this understanding, we come to this slide where I would like to conclude. We saw
few examples on interaction design, such as aviator example, analyst recording the MRI. Such
examples clearly indicate the significance and relevance of human computer interaction.
Another thing which we learned is the software interface design followed by the software
design cycle. And finally, we studied about the user system and characteristics, how we can
strengthen this system and how can we boost its characteristics.

So, I will rest my case here. We will see each other in next lecture. Thank you. Thank you.