User testing is easy. Just ask people to carry out tasks on your app or website. There’s even services who will do it for you in an hour.
But what about when you want to go further and find out how well your site’s content solves your visitors’ problems?
What if your visitors’ problem is that they are deeply concerned about the mental health of a young person they know?
You may have good quality mental health information that’s easy to find and understand but how do you know if it is the right information to solve their problem?
You could test how easy it is to find. You could test if they think it is good quality. But how do you test if it is actually what they need when faced with a real situation?
Well Informed is a website for non-mental health professionals who are worried about the mental health of a young person they know. It’s designed to provide them with quick and detailed information on 9 common areas of mental health. The site is structured around ‘5 Essential Questions’ on each topic. When it launches it will include dynamic support content.
Does Your Content Work in Real Life?
The site is designed for visitors with immediate and pressing feelings of concern for a young person’s mental health. This meant that to be able to test the site’s effectiveness the Well Informed team needed to find a way to replicate this situation and then ask if its content actually improved confidence and knowledge around what to do next.
The tests needed to find a way to mimic Well Informed users being worried about a young person’s mental health.
The tests could have asked people if they felt more knowledgeable after reading the site’s content, but this would not accurately measure if they felt more knowledgeable and confident about dealing with a real situation. To make sure that Well Informed’s content would help in a real life situation the tests needed to replicate real life as closely as possible. Here’s what they did.
1. They Created Scenarios
The team created three scenarios, each of which placed testers in the role of a non-mental health professional with fears about the mental health of a young person they work with. So as to not be too demanding of testers, the scenarios were short. This also made it easier for testers to bring their own experience to the role. Here’s an example scenario: You’re a key worker in a residential setting, helping homeless young people prepare to move into independent living. Over the last couple of weeks one of your young people, a 16 year old girl, Mandy, has been telling you she keeps feeling sick and dizzy. She says she feels scared and panicky even when there doesn’t seem to be any immediate or obvious cause. You think she might be having panic attacks brought on by anxiety. However you’re not sure what to do about it.
2. They Created Content
The team already had early stage content on the site covering the topics of Depression, Anxiety and Eating Problems so they duplicated this onto three test pages. The ‘5 Essential Questions’ that Well Informed structures its content around are:
- What should I know?
- What works?
- How can I help?
- What do young people say?
- Where to go next?
Page length for each of the 3 topics ranged from 800-1200 words.
3. They Created Surveys
The third element of the test involved creating the right follow up questions to find out how effective the content had been. The team used Poll Daddy to craft a 10 question survey for each scenario, using a mix of closed and exploratory questions. The surveys asked testers after reading the page’s content to rate their level of knowledge and how confident they felt supporting the young person in their scenario. Other questions asked them to explain their answers in more depth, what was most and least useful and whether they would use the downloadable pdf of the page in their work.
4. They Made All-in-One Test Pages
The team then placed each scenario and its corresponding survey onto each test page. Having the three elements on one page meant the testers didn’t need to click to a different page at any time, increasing the chances of them completing the survey. The surveys were embedded using embed codes from Poll Daddy. Each page was wrapped with explicit 3 step instructions on what testers needed to do.
5. They Hammered Email Lists And Social Media
Putting together an all-in-one test page is fine. But you still need to get the right people to come and test it. Sixteen25 made use of its list of 800 non-mental health professionals and, in batches, invited them to test the pages. The team also used a number of groups and forums they were already active in on LinkedIn, Facebook and Parenting Websites. They incentivised participation in the testing with a £100 cash prize. Each week, for 3 weeks, the team invited new batches of people from their lists and social media connections to test the three pages. At the end of each week they produced a new iteration of each page, ready for a new batch of testers the following week. In this way the content developed steadily and the team learnt what additional functionality users were likely to use on each page (e.g. pdf content downloads).
The Results
Testing in this way elicited the following results:
- 3-12 tests per page per week
- 75 tests total over the 3 weeks
- 3 site pages iterated to a higher ‘problem-solving standard’ quality of content, with a clear indication of depth and structure for the other 6
The process also yielded further insights into user needs for dynamic content and ways to interact with the site. Testers said they needed support to apply the knowledge in practice. This includes interactive support to help people work out how much or how little intervention to take on, and how to use their new knowledge in specific practice situations. These features will form the basis of the next round of testing. Editors note: If you’d like to be involved in Well Informed’s future testing you can email the team by clicking here.
