Sunteți pe pagina 1din 17

Usability

Stud
Technical computing
y
Matlab vs TKsolver by James

software
Rockwell

Introduction

Included in this white paper is a comparison of the usability of two technical


computing programs: MATLAB and TKsolver.
As the engineering field relies more and more on computer programs to get a
job done it is essential that those going to school learn how to run at least a
few of the powerful programs available. Before computing programs, all
problems had to be solved by hand with a pencil and paper. As computers
became popular and people realized the potential they had, computing
programs were created. These allow an engineer to run millions of
calculations in fractions of a second. Even simulations that would have had
to have been done in the field can now be done with a computer.
MATLAB and TKsolver are very popular programs used and taught at the
University of Idaho in the mechanical engineering program. They are
included in various classes starting with the basics early on and moving up to
advanced calculations in Junior and Senior year.
The test I carried out compares the ease of use of these two programs at a
beginner level. Say, a freshman just being introduced to the program. I
composed the test using seven tasks that an engineering student would do
regularly while using one of these computing program.

Methods

This usability test was completed in the same quiet, study friendly office the
entire time. This is how these programs are used because they require fairly
deep thought. Distractions could make using these programs hard to do.
I used my HP Envy to test both programs. The software is available to all
students attending the University of Idaho online. I used the version that was
offered there because it is the version that the audience will most likely be
using.
I came up with seven features to test on both programs that would be
common for someone using this software. They are listed below.
1.
2.
3.
4.
5.
6.
7.

Basic computation
Computation of a series of equations
Find help within the program
Make a list
Plot a list versus another list
Put comments in work (units, clarifications)
Create a deliverable report

After I had tested a feature in one of the programs I would evaluate it using
the following Likert scale. I filled out a Likert scale for each of the eight
features listed above for both programs. This makes a total of 16 completed
Likert scales.

Strongl

agree

y agree

neutr

disagr

Strong

al

ee

ly
disagr

Tool was easy to access and


use

ee
1

Task was completed in a time

efficient manner
The result was satisfactory

The Likert scale rates each feature in three different categories from a scale
of one to five, five being the best and one being the worst. The total score for
each category in the program will be found by summing the averages from
each scale from each test.

Test Results
Basic Computation
MATLAB
Strongl

agree

y agree

neutr

disagr

Strong

al

ee

ly
disagr

Tool was easy to access and

ee
1

use
Task was completed in a time

Strongl

agree

neutr

disagr

Strong

al

ee

ly

efficient manner
The result was satisfactory
TKsolver

y agree

disagr
Tool was easy to access and

ee
1

use
Task was completed in a time

efficient manner
The result was satisfactory

After the Basic computation test for both programs I was much more
impressed with MATLAB over TKsolver. MATLAB allows the user to simply put
in the problem to be solved with values and press enter. An example of an
input in MATLAB is 2+2. The answer is displayed in the answer box as a
numerical value. This is user friendly because variables dont have to be
used when they arent necessarily needed. TKsolver requires the full
equation to be written out with a variable for an answer. This takes more
time than MATLAB does to do the same simple computation. An example of
an input into TKsolver is 2+2=C. The answer will be displayed as an output
of 4 corresponding to the variable C. Although MATLAB earned a better score
in the basic computation test, both resulted in satisfactory results.
Computation of a Series of Equations
MATLAB
Strongl

agree

y agree

neutr

disagr

Strong

al

ee

ly
disagr

Tool was easy to access and

ee
1

use
Task was completed in a time

efficient manner
The result was satisfactory

Strongl

agree

neutr

disagr

Strong

al

ee

ly

TKsolver

y agree

disagr
Tool was easy to access and

ee
1

use
Task was completed in a time

efficient manner
The result was satisfactory

Solving multiple equations simultaneously is what these programs do best.


They are built to be able to solve the tough problems that engineers come
across regularly. This test, remember, was taken with a beginner audience in
mind. A more experienced MATLAB user might be able to argue that MATLAB
is easier to use than TKsolver for solving multiple equations at once.
However, to the beginner TKsolver is much easier to use. Each equation can
be written in the rules section of the page and the inputs can each be
assigned to their corresponding variable in the variables section. When solve
is clicked all of the equations are solved for and the outputs are displayed

next to their corresponding variables.

The Rules Sheet is the top section where equations are written. The bottom
section is ht evariables section where the inputs and outputs are displayed.

This is very user friendly for the beginner TKsolver user. MATLAB has a longer
process to be able to solve for multiple equations at the same time. A script
must be made and saved before it can be run. This can be confusing for a
new user. The image below shows a script and the output.

The script had to be made, written, and saved into a folder before it could be
solved for. This can be confusing and irritating to someone who isnt familiar
with the program.

Find Help within the Program


MATLAB
Strongl
y agree

agree

neutr

disagr

Strong

al

ee

ly

disagr
Tool was easy to access and

ee
1

use
Task was completed in a time

efficient manner
The result was satisfactory

Strongl

agree

neutr

disagr

Strong

al

ee

ly

TKsolver

y agree

disagr
Tool was easy to access and

ee
1

use
Task was completed in a time

efficient manner
The result was satisfactory

Finding help within both of these programs was not hard. I like that TKsolver
has readily available tutorials with only one click. Tutorials are the most
effective teaching tool for me. MATLAB also had great options for help under
the help drop down. Another valuable resource for MATLAB help is the
worls wide web. There are many online resources for help and direction in
MATLAB. This can be very helpful when trying to work through problems in
code.

Make a List
MATLAB
Strongl

agree

y agree

neutr

disagr

Strong

al

ee

ly
disagr

Tool was easy to access and

ee
1

use
Task was completed in a time

efficient manner
The result was satisfactory

Strongl

agree

neutr

disagr

Strong

al

ee

ly

TKsolver

y agree

disagr
Tool was easy to access and

ee
1

use
Task was completed in a time

efficient manner
The result was satisfactory

Creating a list in these programs is a very common thing. A list is used


whenever a large group of numbers needs to be tested in a row or the user is
in need of a plot. In both of these programs making a list was fairly
straightforward. MATLAB is much quicker than TKsolver when making a

simple list. What makes MATLAB hard to use in this aspect is that you must
know the correct callout for a list. TKsolver makes this very simple. Clicking
list causes a page to pop up that allows the user to name the list. Once
named the user clicks the empty list and manually fills it. This takes much
more time than the one, simple command in MATLAB. For a beginner who
doesnt know the command for a list in MATLAB finding the answer could be
time consuming, but for an experienced user it is much more efficient than
TKsolver.
Below is a side by side comparison of the process of creating a list in MATLAB
and TKsolver.

MATLAB:
left

Plot a list versus another list


MATLAB
Strongl
y agree

agree

neutr

disagr

Strong

al

ee

ly
disagr

Tool was easy to access and

ee
1

use
Task was completed in a time

efficient manner
The result was satisfactory

Strongl

agree

neutr

disagr

Strong

al

ee

ly

TKsolver

y agree

disagr
Tool was easy to access and

ee
1

use
Task was completed in a time

efficient manner
The result was satisfactory

The same test results are true for making a plot in these two programs. If the
user knows the command for creating a plot in MATLAB it is much more
efficient. The problem is that for beginners that might not be known, and
finding the answer can be time consuming as well as frustrating. In TKsolver
making a plot is easily done without any prior knowledge of the program. It
has a step by step path that leads the user to the end. This process takes
longer than MATLAB however which is why it gets a poor efficiency rating.

Put Comments in Work

MATLAB
Strongl

agree

y agree

neutr

disagr

Strong

al

ee

ly
disagr

Tool was easy to access and

ee
1

use
Task was completed in a time

efficient manner
The result was satisfactory

Strongl

agree

neutr

disagr

Strong

al

ee

ly

TKsolver

y agree

disagr
Tool was easy to access and

ee
1

use
Task was completed in a time

efficient manner
The result was satisfactory

Putting comments into code is very important in computer computations. It


helps to keep the user from getting confused, and allows for somebody else
to understand what is going on in the code. Both programs have satisfactory
options for commenting code but TKsolver is more beneficial to the user. It
has space for comments in the rules as well as in the varibales. This allows

the user to identify variables, give reasons for equations, input units, etc.
MATLAB doesnt have the ability to label as thoroughly as TKsolver does.
Below is a comparison of commenting in both programs.

MATLAB: left, commenting in green


TKsolver: right, commenting in far right
column

Create a Deliverable Report


MATLAB
Strongl

agree

y agree

neutr

disagr

Strong

al

ee

ly
disagr

Tool was easy to access and

ee
1

use
Task was completed in a time

efficient manner
The result was satisfactory

TKsolver
Strongl

agree

y agree

neutr

disagr

Strong

al

ee

ly
disagr

Tool was easy to access and

ee
1

use
Task was completed in a time

efficient manner
The result was satisfactory

The version of MATLAB that is offered to students at the University of Idaho


lacks a report option. After code is written and plots are made there is no
way to create a summary of all your work in MATLAB. The code must be
copied into word and printed. This is inefficient and frustrating. TKsolver
offers a report option that puts all of the data form your computations into a
report that can be easily manipulated to exactly what is wanted. This is a
very useful tool to students that are taking classes on these programs. It is a
very well put together written deliverable. Below is an example of part of a
report created in TKsolver.

TKsolver gives the user an


option of creating a report.
This report is very easily
manipulated.

Conclusion
The average score in each category are as shown below. .
MATLAB
averag
Tool was easy to access and

e
3.7

use
Task was completed in a time

4.1

efficient manner
The result was satisfactory

4.3

TKsolver
averag
Tool was easy to access and

e
4.9

use
Task was completed in a time

3.9

efficient manner
The result was satisfactory

The test I did proved that for the chosen audience of beginners TKsolver is
more usable than MATLAB. These results could be completely different for a
different chosen audience. For a beginning user that wants to use a

computing software TKsolver is an easy to use and effective program.


MATLAB has a steeper learning curve and requires more knowledge of
commands etc. which makes it harder for a beginner to use. Another big set
back for MATLAB is that it doesnt offer a report option. Having an easy way
to create a deliverable report is a very usable tool. I would recommend for
MATLAB to offer this tool.

Self Evaluation:
I think that I did and adequate job of testing these two computing programs.
This project gets very monotonous but I enjoyed looking at the end results. I
am a little biased towards TKsolver because I have had such a hard time
learning how to use MATLAB efficiently. This may have skewed my results a
tiny bit but I tried to stay as unbiased as I could. This type of test seems like
a very useful tool for anybody that is building anything, Asking the consumer
to evaluate a product gives excellent feedback to the producer.

S-ar putea să vă placă și