Human Factors in System Design: Usability Study of mHealth Diabetic Self-Management Application

This report contains a usability study of cognitive walkthrough stating the user-centered cognitive walkthrough that is capable of addressing all deficiencies contained in the original technique and perform validation with TA protocol (Think aloud). The study evaluates the effectiveness, user acceptance, and efficiency of the mHealth Diabetic self-management application. The report includes the interactive system and its users, use cases, usability requirements, evaluation methodology, evaluation, and findings of evaluation.

Running head: HUMAN FACTORS IN SYSTEM DESIGN
Human Factors in System Design
Name of Student-
Name of University-
Author’s Note-

1HUMAN FACTORS IN SYSTEM DESIGN
Table of Contents
Part One: Interactive System and its Users......................................................................................1
Part Two: Use Cases........................................................................................................................1
Part Three: Usability Requirements.................................................................................................2
Part Four: Evaluation Methodology................................................................................................3
Part Five: Evaluation.......................................................................................................................4
Part Six: Findings of Evaluation....................................................................................................10
Bibliography..................................................................................................................................13
Appendix........................................................................................................................................14

2HUMAN FACTORS IN SYSTEM DESIGN
Part One: Interactive System and its Users
Interactive system that is being chosen for carrying out the evaluation study includes a
walkthrough for evaluating mHealth Diabetic self-management application. This evaluation will
increase the usability and the clinical efficacy for the mobile application. The type 2 diabetes
affects more than 382 million people all over the world. This particular number is expected to be
increased by approx. 35% in coming years. Managing the chronic diseases by own using the
mobile application has become common. The application of mobile health is mainly delivered by
smart phones or by tablets providing the patients potential to manage their own diabetic rate.
The mobile applications are new technologies and there are few studies that have tested
those applications as clinical interventions. The researchers stated that the people who are using
the diabetes application via their smartphones can use it for tracking the blood glucose and the
diet helps in controlling the adherence for managing the diabetes and improves self-monitoring.
The evidence on the diabetic app of clinical effectiveness is very inconclusive. The methods
including the usability testing is costly and is also time consuming. This report contain the
usability study of cognitive walkthrough stating the user centered cognitive walkthrough that is
capable of addressing all deficiencies contained in original technique and perform validation
with TA protocol (Think aloud). This approach will assess the effectiveness, user acceptance and
the efficiency with the diabetic patients on application of self-management.
Part Two: Use Cases
The think aloud (TA) protocol that is used for this self-evaluation method includes fields
of human factors as well as computer engineering. In this evaluation, the users includes some
representative tasks where the thoughts are verbalized in their interactions. The aim is gaining

3HUMAN FACTORS IN SYSTEM DESIGN
understanding through the task behavior of the users, the thought process of the users for
conducting system interaction and helps in identifying usability problems in system. The user
who are involved in the evaluation are the actual system user. The walkthrough is that is carried
out in this evaluation includes 6 to 10 developers, 2 to 3 engineers and 6 to 10 system users who
are associated with the group session. The group includes many tasks as well as participants who
are included in the group session.
Part Three: Usability Requirements
All the users associated with the system includes the group tasks uses and the participants
take part of the intended users. The users involved expresses the views prior other groups start
determining all problems in the self-evaluation system. The technique that is added in this
technique includes performing conventional expert of the session and added video recording for
the cognitive walkthrough to show the users interactions with this system. This usability study
includes think aloud session with all users for getting the conventional walkthrough of the
experts.
The tasks included in evaluation are much important and those are evaluated to find the
issues and the system limitations. The tasks are to address the area involved in the system and
includes sufficient number as well as specifications. In this analysis there is limited tasks
analysis and all the tasks are broken into small actions. There is no proper guidelines for the
process of task development and for task selection and also helps in priotirizing the system or the
tasks included in evaluation.
The study that is undertaken for this evaluation is mainly conducted at University of Utah
in the Salt Lake City of USA. After receiving the approval of Institutional Review Board, 12

4HUMAN FACTORS IN SYSTEM DESIGN
number of patients that were recruited. Setting that was made included the diabetes and the
endocrinology center for the University of the Utah Health. Cognitive walkthrough was mainly
conducted in conference room by using a projector as well as all individual think aloud session
was done in a quiet place in clinic. All the participants participated in the evaluation test has to
fill out the informed consent at starting of the evaluation session.
The 12patients who were involved in the evaluation criteria were recruited in the
Diabetes and Endocrinology Center using the purposive sampling for ensuring users having
different characteristics representing method evaluations. Criteria inclusion included were a)
diagnosing the patients for the Type 1 and the Type 2 diabetes, b) having no cognitive
impairment, c) familiarity as well as knowledge and knowledge regarding using computer
system, internet as well as mobile phone, d) should have the ability for speaking as well as
understanding English language. The patients included should have no connection with the
mHealth application that is being evaluated. The patients are to be randomized in two different
test groups having 6 participants in each group for performing the user centered cognitive
walkthrough. After the completion of the test, the participants were given $20 gift cards to
participate in the evaluation.
Part Four: Evaluation Methodology
There were suitable scenarios as well as tasks that were developed as well as validated
for carrying out this evaluation. The tasks were mainly based on some extensive review for the
diabetes guidelines that comes under self-management. This would ensure that all important as
well as self-management aspects that are critical for the evaluation is included in the tasks. The
tasks that were included in the system were then reviewed by three domain experts. The tasks

5HUMAN FACTORS IN SYSTEM DESIGN
were also reviewed by the diabetes patients to ensure accuracy and were accessed with the use of
content validity index. This validation iterations that were included in the evaluation test before
reaching the acceptable score were being completed and the result of 14 tasks are listed in the
appendix 1.
Evaluation involved in informing the participants in the evaluation test were done for five
minutes and the application of the mHealth were then demonstrated to the participants on large
screen for about 15 minutes. All the patients associated with the study was provided with a
booklet having 14 tasks. All the scenarios and the tasks included space for recording individual
observations as well as provide answers for the two simplified cognitive walkthrough questions.
This is done for assisting the group discussion in the evaluation study. In the group walkthrough,
the participants involved read out all the scenarios in turn and the group for evaluation study
guided facilitator to perform all the tasks that are shown on prjector screen. After performing
each tasks, the patients expresses individual personal thoughts that they experienced about the
usability problem for mHealth Application undertaken for the study.
The application of mHealth is a platform that is used to research to support self-
management for the Type 1 and the Type 2 diabetic patient. This application is considered much
suitable for the study by some researchers that leads to some system improvements. The mHealth
application mainly includes different diabetes functions that allows the patients for monitoring
the blood glucose, the medication, food as well as physical activity included in the system. The
measurements of blood glucose gets transferred to a glucometer through Bluetooth are even
entered manually. The food as well as the physical activity data are noted manually. The glucose
measurements are mainly are shown on graphs.
Part Five: Evaluation

6HUMAN FACTORS IN SYSTEM DESIGN
Usability problem characteristics: Usability problems included were not small. In the
Main View, it is not possible to figure or identify the ranges that were based on the own need of
the patients for measuring their latest glucose level. The most important feature that is
considered is individual tailoring of diseases. The second people included in the evaluation study
is the problem involved in Main View. The navigational concern involved in Main View was the
most important concern and the participant needs to swipe across graphical image for navigating
the periodical pattern for checking the glucose measurement of the patient for the last group.
This included different action compared to previous steps included in the evaluation.
There is another personalization problem that is plotted in Plotted Graph View. This
consists graphs that includes cut-off that shows low values, middle value as well as high value of
glucose that are included for all the individual patient. In the List View, the patients found that it
was not clear about the way to navigate as well as define the location when the users came in the
list in the diary of the registered food, for medication, for insulin values, as well as glucose. It is
not possible to get the response for the date sequence for all the entries listed and that lead to
confuse the patients.
Time estimates for evaluation: Elapsed time included for evaluation included in entire
UC-CW evaluation process was taken to be 1830 minutes. Most part of time was elapsed for
reviewing all guidelines as well as constructing the evaluation tasks and the data coding along
with the usability problem that is determined for the each task. Less time was consumed in
validating the tasks. About 480 minute was elapsed for problem determination and 130 minutes
was taken for task validation in the system. Assigning heuristics as well as the severity ratings
took about 120 minutes. The user evaluation for the evaluation technique took about 140 minutes
for completing the evaluation. The total time of evaluation is mainly divided in to 26 problems

7HUMAN FACTORS IN SYSTEM DESIGN
that are identified and that is equal to 70.4 minutes for each usability problem. Session of
evaluation took about 14 minutes and this included 5 minutes for informing method and 15
minutes for application tutorial that is given to the participants. 120 minutes was taken for
evaluation.
The Main View page includes the UC-CW and think aloud included had the highest
quantity of usability problems. This was mainly followed by Carbohydrate entry view. This is
shown in the figure below. Both methods that is included in List View method, Glucose entry
method and includes time setting that has very low volumes for the usability problems as well as
the heuristics violations identified in the application. The main view and the time setting view
has same severity rating. This is similar to the plotted graph and the list view graph.
Figure 1: Usability problems and number of heuristic violated
(Source: Georgsson et al. 2019)

8HUMAN FACTORS IN SYSTEM DESIGN
Total 74 heuristic violations occurred or views in UC-CW with 51 heuristics violation sin
TA. In the figure 2 below, it can be seen that the majority heuristic violation includes both the
methods that stated the category match with the system and the real world. The consistency as
well as the standards at analyzed at 22.9 % and 29.4 % and the recognition or the recall was
analyzed at 17.5 % and with 19.6 %. The user control included in the evaluation and the freedom
has very few violations included by both the methods across the view and that is measured to be
at 1.4 % and 3.9 %. This is except the categories Help and Recover. This does not have any
match category.
Figure 2: Heuristics violation across methods
(Source: Georgsson et al. 2019)
The UC-CW mainly resulted in total 26 usability problems while think aloud it resulted
in 20 usability problems. The UC-CW was mainly identified with 5 catastrophic and include 10
severity ratings whereas in TA two catastrophic as well as 8 major ratings were studied. TA

9HUMAN FACTORS IN SYSTEM DESIGN
included 10 small ratings as well as included UC-CW nine and two cosmetic rating were given in
the UC-CW.
Figure 3: Usability problems, and classifications on heuristics
(Source: Georgsson et al. 2019)
For all individual methods, there are similar results. Both of them are identified with the
most critical problems in the Main view. There were less usability problems located on main
view of the system. Three view were identified, one is glucose entry view, second is periodical
pattern graph, and the third is plotted graph view. This also included one minor rating in each
evaluation tasks and carbohydrate view included minor ratings as well as shows the plotted graph
view. The list view had one small rating with each TA.
UC-CW that were identified includes many ratings that were severe including same
situation list view. The methods included in the evaluation also varied with the severity ratings
for Glucose Entry view, this includes the carbohydrate view in the analysis as well as periodical

10HUMAN FACTORS IN SYSTEM DESIGN
pattern view. This helps in demonstrating the UC-CW includes more sensitive with the severity
ratings for all the views. Both the methods included same severity ratings available for the List
View, for the time setting view and available for the main view. Both the methods has same
average severity ratings included in the chosen application.
The UC-CW that is consumed in this evaluation includes total time and the result
includes in identifying all usability problems. This includes 70.4 minutes that were included as
per identifying the usability problems. TA in the analysis consumed much less time that is 80.2
minutes for the usability problems in the system. The UC-CW was 10 minutes faster as per the as
per the problems that were identified in the evaluation procedure. There are different components
include din evaluation procedure and in the evaluation process which is less time consuming
compared to TA. This evaluation mainly included evacuation session including 140 minutes for
the UC-CW method as well as includes about 281 minutes for TA method. The data coding and
the included problem usability includes the determination consumed about 480 minutes in UC-
CW method and took about 724 minutes for the TA method. For conducting the evaluation
process, the method of UC-CW consumed a total of 5 minutes as well as for the TA method, the
total time consumed for evaluation is 70 minutes. The below figure 4 shows the details of
evaluation method and the time taken for each evaluation in both the methods.

11HUMAN FACTORS IN SYSTEM DESIGN
Figure 4: Time in minutes included in evaluation procedure
(Source: Georgsson et al. 2019)
Part Six: Findings of Evaluation
With the enhanced cognitive walkthrough, the user centered was designed for including
the users in all the phases. This helps to streamline the development of tasks as well as includes
different evaluation process as well as has mportant tasks with the validity aspects for detecting
the usability problems that are critical for all the users. For performing the beginning validation
against original think aloud with small group of user patient measuring the effectiveness, the
efficiency and the user acceptance of the application. Both of the methods includes two domain

12HUMAN FACTORS IN SYSTEM DESIGN
facilitator, includes capturing of digital data as includes structured data analysis for facilitating
validation.
The two groups of people involved in the evaluation has similar demographics, same
technological usage, and related knowledge of mobile phones and includes views related with the
technology having different minor exceptions. This exceptions includes different personal
characteristics such as education, internet knowledge, computer knowledge and their personal
experience. The think aloud group has more negative view on the possibility of the technology in
the health care. These variables were considered as main variables that were sampled in this
evaluation. There are many other aspects such as information and health literacy, the
socioeconomic levels, and includes low levels of education that are assessed for future studies
having some large group of participants.
Usability problems identified: The method of evaluation includes several different and
interesting results. Total number and severity included in identified usability problems along
with the user centered CW were high compared with the think aloud. The evaluation identified
deficiencies that is needed to personalize the features that includes data ranges is basically not
configured for low glucose rate, medium glucose rate as well as high glucose rate in main view.
All the users cannot configure the delimiter range using this application. This application use a
graph that is based on the cut off levels of the patient. The most importance to individualize the
mHealth interventions should have similar needs so that it is individually adaptable for some
specific needs. In the UC-CW method only, the deficiencies were identified. The critical
problems that were identified in group portion included higher discussion level that the issues got
discovered in group portion for the session that includes high level discussion containing Type 1
as well as Type 2 patients. The evaluation also states how mHealth application are applied for

13HUMAN FACTORS IN SYSTEM DESIGN
their needs. Both the evaluation having same severity ratings for identifying the problems and
includes heuristic categories of violation.
Efficiency including time elapsed for methods: For streamlining the traditional and the
time consuming steps, the CW is to be improved. The steps that are required includes tasks
development was modified for including some important measures of validation. The results that
were viewed includes adding validation with respect to UC-CW. This method is more time
consuming compared to the TA method. There are many usability problems identified in UC-
CW as well as include critical problems and this method was more efficient compared to TA
method and is more effective with the time that is included in the usability problem. The
evaluation method carried out under UC-CW was faster than the other method. The portion of
evaluation was also fast with this method.

14HUMAN FACTORS IN SYSTEM DESIGN
Bibliography
Georgsson, M., Staggers, N., Årsand, E. and Kushniruk, A., 2019. Employing a user-centered
cognitive walkthrough to evaluate a mHealth diabetes self-management application: A case
study and beginning method validation. Journal of biomedical informatics, 91, p.103110.
Orfanou, K., Tselios, N. and Katsanos, C., 2015. Perceived usability evaluation of learning
management systems: Empirical evaluation of the System Usability Scale. International Review
of Research in Open and Distributed Learning, 16(2), pp.227-246.
Yousef, A.M.F., Chatti, M.A., Schroeder, U. and Wosnitza, M., 2015. A usability evaluation of a
blended MOOC environment: An experimental case study. The International Review of
Research in Open and Distributed Learning, 16(2).
Paz, F. and Pow-Sang, J.A., 2016. A systematic mapping review of usability evaluation methods
for software development process. International Journal of Software Engineering and Its
Applications, 10(1), pp.165-178.
Toribio-Guzmán, J.M., García-Holgado, A., Pérez, F.S., García-Peñalvo, F.J. and Martín, M.F.,
2017. Usability evaluation of a private social network on mental health for relatives. Journal of
medical systems, 41(9), p.137.

15HUMAN FACTORS IN SYSTEM DESIGN
Appendix
Appendix A
Appendix B