Evaluation of Project

I have found that this whole process is key to success with any product design. In my case, it is extremely useful in building a user-friendly website that works for its target audience. A successful website makes a successful business in this day and age. Companies are becoming more and more reliable on the web because it is so accessible to almost everyone. It widens up the market dramatically and can drastically change a company's income. Websites are the new way to advertise, buy and sell products, but you won't do very well if you don't consider your target audience. Every company needs to be aware of the people their aiming their products at, to get their thoughts and feelings. The user-centred design process gives you steps to follow to get the product that your target audience will want.
A huge part of this process is testing. It is vital that testing is done so you know exactly what your users want. Working with a website, it is vital that it is tested, especially for navigational purposes. The designer may think it's obvious on what buttons go to what page because they've been working on it for some time and are used to it. To twenty different people, it might not be so obvious. Simple methods like paper prototyping can be used to test the navigation of a site. A site that doesn't make its users think about what they're doing is a site that works. If you get no complaints and positive feedback, then you're on the right track. Testing straight away will save you so much time and money in the long run, instead of paying to have a site made and then find out that people find it stressful to use because they can't navigate around it and then pay more money to have it re-designed. You can also get feedback on the look and feel of your product. Colours, fonts, layout and imagery are equally as important to have an aesthetically pleasing product. Simple questionnaires can easily give you the feedback and criticism you need on the look of a design.
Another greatly important task is analysing and evaluating research and test results. It's all well and good that research and tests have been conducted, but it's what you do with those findings that matters. Fitting the design of the product around the information you've found out is important, so it meets the requirements.
For this project, keeping a blog has been very useful to make note of what I've done as I've gone through this process. It's great for keeping record so I don't have to try and remember what happened during the tests and write up everything at the end.

My Designs

Here are the designs I made to solve the milk problem, which was the only one made, so no other midifications were needed:I also experimented with the look of the home page:

Design Specification

The design specification is a very important document in the design process as it contains all the information necessary for a designer/design team to produce a solution to the design problem and to show to a client. It splits the problem up into smaller categories to make it easier to look at. The document should list all the requirements that the product must meet, with any constraints that may affect the product. The customer (Sainsbury's in this case) should be consulted as fully as possible while the document is being made because their requirements are very important and are what the designer should be trying to deliver .

Screen Captured Usability Test

Unfortunately I couldn't get the last screen captured test to upload properly, so I did another test on myself, which I know isn't entirely fair because I've seen the site before, but it's just an example of what you would have seen if the other video had worked. Using the Camtasia Studio software you can follow my mouse and see where I click, as well as see my face and hear my voice describing what I'm doing. I shopped for three items: milk, bread and spaghetti, then checked out using false details.

Usability Test On Final Design

What would normally happen after the task analysis and the testing on the current design, would be an evaluation on the results and then a re-design if appropriate. The new design would then be tested as well, but in my case, I can't re-make a working prototype on the scale of the Sainsbury's website in the given time for this project. Later on I will be showing some of my own designs that would solve the issues found in the testing. What I've done here is tested the site as it is again, but adjusting the the task slightly and adding some questions to the de-briefing questionnaire. You can see from the images below the changes made. The script was kept the same.




















































































Here is the video recording of the test:

Usability Testing from michael bale on Vimeo.

(There was a problem uploading the screen capture recording. You can see what happened from the other video, but I will try get this one working.)

This user had no real problems using the site. Having never used the site before, they found the navigation layout was very helpful. There was one slight fault, which was down to the test organisers. The user thought that they had to log in first using the false details provided, but they wouldn't work because because those details hadn't been registered yet. This was the first time that this has happened, but it should have been made clear to the user that the false details were for after checking out.

Coming next in the next few posts will be a design specification document and my own design ideas of how the supermarket website could be layed out.

Paper Proto-typing

Paper proto-typing may seem a little childish, back to using glue and sticky tape like creating some sort of 'Art Attack'. However, it is a low cost and quick way of testing the navigation and basic function of your site so you can spot any problems and solve them early on in the design process.

Using the online shopping theme, a group of us created a paper prototype of a website and gave testers the scenario of buying some milk.

Here is the network plan we came up with:

We then created five pages which we were going to be using:

Homepage:

Drinks page:

Groceries page:
Dairy page:

Milk page:

We also had a check out page which thanked the user, a loading page if the computer had to add any additional things to the page and an under construction page if the tester selected a page we hadn't created and wasn't necessary for this scenario. As you will soon see from the recordings, the user was asked to navigate through the proto-type using a pen as the mouse, to buy 8 pints of semi-skimmed milk.


User A



User B



User C




As you can see, the milk was again a problem, like it was in the Sainsbury's site. People keep searching for it in the drinks section, so to solve that problem, the item should be added there as well as being in the dairy section. It will reduce shopping time and increase customer satisfaction. They won't get frustrated because they can't find what they're lookibg for. This could result in them using the website again and perhaps telling others how nice it is to use, therefor boosting the supermarkets profits.

Test Results

Here are the signed consent forms:


Here are the filled in questionnaires:


Next I will be showing the recorded results:

User A




User B




User C (has three parts due to camera problems)








From these results, I can work an average of how long it takes people to do a little bit of online shopping using the Sainsbury's website. I noticed that with each user, they spent the longest searching for milk and they all comented that they had problems finding the correct categories for certain items. My next post will be on paper prototyping and its uses.

What To Do While Testing

Test users will feel under pressure. Knowing that they are being observed and recorded increases anxiety. Test users will make errors.

Here's some useful things to consider which I picked up from looking at Rubin & Chismell's Handbook Of Usability Testing and Usability Engineering by Jakob Nielsen:

• Make test users feel as comfortable as possible.
• Make clear to users it is the system being measured, not them.
• Make clear to users no information will be revealed - Everything is confidential.
• All test users should remain anonymous.
• Inform them they can stop the test at any time.
• Minimise the number of observers.
• Ensure you've answered all their questions before you start the test.

How to achieve this:

• Environment should be comfortable/quiet
• Serve refreshments
• Make sure everything is ready before hand i.e. computer, hand-outs, recording devices...
• Turn mobile phones off
• Make the first task relatively easy
• Make sure there are no disturbances

Testing With Real People

It's time to prepare what I need for testing an existing website (Sainsbury's) with real people. It's not enough to show people pictures of a website, because you won't get what they think of the usability, only on it's appearance. So here's what I will need to do to carry out a test with real users and a working site:

• A consent form to have the user's permission to use the recordings from the testing. Here's what I'll be using:

Participant Consent Form

User-Centered Design – Task Analysis

By signing this form I agree to participate in this research and understand:

• This agreement is of my own free will

• I have had the opportunity to ask any questions about the study

• I realise that I may withdraw from the study at any time without giving a reason

• I have been given full information regarding the aims of the research

• All personal information provided by myself will remain confidential and no information that identifies me will be made publically available

Signed: ................................................. Date: ....................................

(by participant)

Print name: ..................................................................

Signed on behalf of researcher

Signed: ................................................. Date: ....................................

Print Name: ..................................................................

• A background questionnaire - So when looking at the results you can deduct why the results turned out a certain way by looking at the person's backround information. This is the questionnaire I will be giving my users:

Age:__________

Gender: M / F

Income:

0 – 15K __________

15 – 25K __________

25 – 40K __________

40K + __________

Computer skills:

Terrible	Not Very Good	Good	Very Good

How Often Do You Shop Online?

Never	Hardly	Once A Month	Once A Week	Everyday

Do You Shop At Online Supermarkets? Yes / No

If Yes, Do You Shop With The Online Supermarkets That Do Delivery? Yes / No

• A script of what I should read to each user before doing the test, which explains what they need to do. A script is used so each user is given exactly the same information. This is my script:

Thank you for agreeing to participate in this usability study today. I will be working from this script to ensure that my instructions to everyone who participates in the study are the same.
The aim of the study is to observe how people shop online using a supermarket website.
This session will be recorded so I have a back-up of how each study went and so I can look back at them for further analysing.
In this study you will be given a simple task of buying a few items on a supermarket website. This isn't a test on you, it is a test on how easy the website is to use.
You will start the task from the website's homepage. As you work through the task, say aloud what you are doing or thinking. For example, if you are clicking a button, say which one you are clicking and why. e.g. "I'm looking for bread so I'm clicking on the bakery button."
At the end of the task I will ask you some questions about the website.
You've read through the task and the information that is needed for it. Are there any questions before we begin?

• The task problem. I will be giving the user a sheet of what they need to do and any information they will need to use to do it. Here is what it will say:

Task Analysis Problem

You would like to buy some items from an online supermarket site and have them delivered to your home.

The 5 items you require are:

1. Bread (wholemeal, unsliced);

2. Milk (semi-skimmed, 2 pints, organic);

3. Red wine (price limit £5.00);

4. Spaghetti 500g (price check);

5. Tin tomatoes (chopped).

I will also supply the following details, in order for the user to get up to the purchasing stage, of which need to be different for each test user:

• Name

• Address

• Card Details

• Video/audio device to record the session and something to make notes with
• Debrief questionnaire. Questions will be asked verbally and this session will also be recorded. Here are the questions I will be asking:

Task Analysis Questionnaire

1) What were your general first impressions of the site?

2) How easy did you find it to locate the items on the shopping list?

3) Were there any particular problems you encountered when using the site?

4) Were there any features of the site that you found difficult to use?

5) Were there any features of the site that behaved in unexpected ways?

6) Were there any features you would like to have seen on the site that were not available?

7) How do you think the site could be improved?

8) Would you use the site again?

• Then I will thank the user and say goodbye.