Navigation

Perfect your website with 3 simple UX research methods

By Paula Borowska | UX Design | Apr 18, 2014

User testing doesn’t have to be difficult at all. I’m going to talk about three lean research methods that are neither extensive nor expensive to carry out but will in turn provide you with valuable feedback.

The methods can be used for anything, test a landing page or test a whole mobile app. The important thing is to gather feedback: design for yourself, and only your point of view counts; design for others, and you need to hear their thoughts.

Which method, or methods, you choose will have significant consequences for your design, and different methods are more suitable for different objectives.

 

Experience sampling

This technique is used to ask participants a single question throughout their day to note their experiences. Keep it very simple like “What events do you input into your calendar app?” The idea is to get a number of people answering at different times of the day to yield various responses. You then collect the answers into a spreadsheet and analyze.

Results

The results provide insight about what the user needs, identify pain points and delight, currently in your product. It can bring inspiration about new features—things currently missing in your current version.

Preparations

You’ll need to figure out a bunch of things in order to run the test. First, what question are you trying to find the answer to? How many times in the day will you ask it? How you’ll you be asking—email, text message, phone call, online survey? You need to decide how to collect this data. You’ll also need to figure out who will be in charge of collecting this data and analyzing it.

You will also need to figure out who the participants are, how many of them will be in your study? Make sure that they can receive your question—not everyone has access to iMessage, so you can’t send it over from your computer. When it comes to the participants, you’ll need to make sure they are prepped themselves. Go over the test expectations, ie. the question will be asked 4 times through out the day. It’s also a good idea to run a practice test to make sure the participants can follow along throughout the test. This will wrinkle out any concerns or questions about it.

Running the test

It’s great to let the participants know they are doing well. You should be monitoring the answers as they come in and let the participants know if they have not answered the question correctly, they may have misread or misunderstood it. And of course, on conclusion thank them for their time.

 

Card sorting

This research method is meant to uncover the best way to structure information within your design. In the test, participants will group items into a sequence that is logical to them. If you are not sure about the navigation of your site, or the categories for your app, this test is perfect for you.

Results

What you can expect from this test is to figure out the way your target audience perceives your content. You’ll have a new way to think about terminology, the relationships between the items, and any missing features; all thanks to a dendrogram structure resulting from the test.

Preparations

This test can run in so many different ways! You’ll need to decide what content you want the participants to test, it can be current content or future content you’d like to create. You should not have more then 100 items to sort through; it can be quiet complex to sort things so the golden rule is less than 100.

There are various ways in which the participants can partake in a card sorting test; you can organize it in person with index cards or create the test online and send out a link. If you do this online, analyzing results can be much easier. Some well know online tools include SimpleCardSort, Optimal Workshop, and UX Punk.

There is more to consider: do you want the participants to preform the test individually or as a group of 3-4 people? Is the card sort going to be open or close sort? (Open sort means that you allow the participants to create and name the groups themselves, as many as they see if. A close sort means that you provide the categories and they’ll have to fit the items into them.)

Running the test

This only works if you are conducting the test in person, but you need to take notes. Write down when the users are confused about an item; if they are unsure help clarify, but don’t lead them with your own suggestions just explain what the item is—make a note that they were confused. If they seem to have trouble placing an item into a group where you see it jumping from one category to another, make a note of it.

 

Usability testing

This is a research method that gives you insight about how users use your product by observing their behavior. This testing is usually done one-on-one where you observe the users’ actions and listen to them comment as they go through the test.

Results

Usability testing points out the flaws and delights in your products. You’ll be able to take away things that currently work well, areas to improve upon, data about satisfaction, currently missing features and user input. It’s much more convincing to bring a quote from a user about a frustration point to your boss than just to say that the frustration point exists. Users are such a powerful motivator!

Preparations

A usability test is a walk-through your site. It usually tests a small portion, like the sign up process or maybe using the core feature of your site for the first time; if you were designing Facebook, it could be the process of signing up and adding a few friends in a couple of different ways.

You’ll need to figure out a bunch more for this test then the previous two. First, what are you testing? What are the steps you want the participants to take? For example, ‘You want to add a friend by email.’ You should also have a bunch of warm up questions just to get the participant familiar.

Additionally, you will be to figure out who will be taking the participant through the test and who will be taking notes. It’s impossible for one person to do both, as notes will be missed. It’s also recommended to record the session on top of that. Allow enough time for one participant to go through the test without rushing them before the next appointment is up; ideal test time is 30 min, so allow for a 45 min session.

Running the test

You should also provide a quick brief to the participant telling them why are they here and what they will be doing, you need to make them as comfortable as you can because being observed is not easy for most people. While the participants are going through the scenarios, insist they think out loud, that’s the only way you’ll be able to get access to their reaction, impression and thoughts; observing is not actually enough. Lastly, don’t help them out! Do not lead them; allow them to figure it out. It’s okay if they fail, or keep on failing. The point of the test is for you to find out where the participants trip up and not to get them from point A to B by holding their hand.

Share this post
Comments (no login required)
  • http://blog.coffeedesign.net/ Ahsan Idrisi

    I never knew that cart sorting was possible online, i had heard this technique..Good to know sites like Optimal Workshop

  • Harish

    thanks nice post

  • Nils

    Can Trello be used for online card sorting? There are no built in analysing tools…