User testing is vital during the research, design ad production process as I need to make sure the project caters for my target audience and is easy to use/understand. Here I recapped on some background research using several resources such as Steve Krug's book on user testing.
- Test a product whether it is your own or not (competitors website to see what works and what does not)
- Test consist of tasks to be completed (observer sees how the participant completes the task here they can see if the function is user friendly or not)
- Observers is there to take notes of occurrences and ask any questions if needed depends on the type pf study)
- Two forms of data types that are used within user testing
1. Quantitative
- This method is similar to scientific experiments as they are working with a hypothesis
- The format uses numbers to gather data instead of detailed information
- Once the sample of questions/tasks has been designed it cannot be changed
- This type of research requires a large sample of participants
- Those taking part have to reflect the actual users that will be using the product once it is complete or up and running
- There should be a very small amount of interaction between the observer and participant
2. Qualitative
- Method used to gather information on how to improve a product regarding it's functionality
- Easily detect errors within the system as there will always be errors and areas to improve on
- The is a more informal approve unlike the one mentioned above
- This is not a scientific method, allowing interaction between the observer and participant if needs be (test can be altered during)
- Only a small sample is need therefore it is easy to recruit participants once the target audience has been detected
- The study can be any length due to the small number of participants that are need
- A debrief session is carried after the tests have been completed, here the problems gathered are discussed and possible solutions are made
- Easy to detect errors even using a small sample
- Observing the participant helps us to understand how they navigate through a website this is helpful as then I can develop my website accordingly
Carrying out the Usability Tests
The usability tests need to be carried out on a frequent basis especially when my design and production process come to place as I need to make sure that all navigation systems are working including the database and cloud functionality. I have decided to stick to a sample of 3 participants as it is easy to recruit and is a good number to find errors within the system. The small number makes it easier to carry out longer tasks and allow the ability to debrief after. Before carrying out a usability test I need to decide on what I am actually test, what I want to find out and why? In order to do so I will need to write up a script where i introduce myself, the project and what I am testing. Within the script \i need to encourage the participant to be as honest and open with their opinions as possible. The tasks will be set in scenarios in order to make it easier to understand.
Napkin User Testing
I will start my design development using napkin test, here several different design sketches will develop and tested. The use of various styles and techniques will allow users to depict one which they prefer or the feature that they prefer. The results provide will determine the layout or feature that I will be developing on. I have chosen to use this method as it is easy and quick (can be done under 5 minutes) to undertake but still providing me with enough information to continue.
Wire frames
The following form of user testing that I will implement after the napkin tests will be the wire-frames. This will consist of creating a diagram of the website where menus, logins and functions will be tested. It is mainly to see if the placement of attributes are effective or not if so why?
After these 2 forms of testing are complete the next stage is to design and test and actual web page although this process will come in later once the design and production process is under-way.
Tasks
It is important that I test only what is need at that particular testing session because of this it is essential to create a list of what I want or need to find out first.
- important tasks (tasks that will always need to be carried out when the user visits the page
- do not allow them to use the search bar for their tasks unless you as testing the functionality of the search bar
- Create a script with the tasks they need to complete, develop them into scenarios as participants tend to work better
- Carry a pilot test before the actual user testing just in case amendments need to be made, it is a quick process allowing you to use anyone as your pilot participant (they don't have to reflect your target audience)
After any form of testing it is vital that I debrief all the errors found (Steve Krug's book recommends a free lunch but as a student this process will not take place). The best way is to list all the problems found then put them in order of how complicated they are to solve. I then need to decide what will be sorted before the next user testing, the book recommends that small errors are dealt with first. All these tests need to be recorded on tape so that it can be looked back upon and used if needed.
Although it is likely that I will encounter on many errors, I need to make sure that they are manageable in terms of fixing. I might find that due to my time frame I cannot fully fix the problem but find a way around it that will not affect the functionality as much.
- Alter features
- Change functions around
The next step is to develop a script with tasks based on the Universities Moodle website, my aim is to find out how students use the system (what is normally done first?) and any constrains that are found within he website.
How to analyze the data
gathered
It is important that once the data is gathered
the results are analyzed, in order to do so it is vital to choose an
appropriate method. I have looked into a variety of forms that are displayed
below:
- Regression:
It looked at the correlation found between both the independent variable and
the dependent variable.
- Correlation:
Looks into the relationship between variables
- Cross
Tabulation: Method is used to find correlations between variables at the same
time; the results are then divided into groups and analyzed.
- Descriptive:
Charts are used to displayed results, as the data mainly gathered is
quantitative. The researcher will also measure the mean, median and mode
Sources
Krug, S. 2010. Rocket Surgery Made Easy- The Do It Yourself Guide to Finding and Fixing Usability Problems. Berkeley: New Riders.
Fink, A. 2009. How to Conduct Surveys: a step by step guide. 4th ed. California: Sage.
Statpac. n.d. Correlation Types. [online] Available at: http://www.statpac.com/statistics-calculator/correlation-regression.htm [Accessed: 17 Nov 2013].