Summary of user research

[calvin-lau-sig7]

A collaborative team from the Ministry of Justice, Department of Work and Pensions and the Scottish government proposed the component after having worked on components in their own services. This slide shows analysis of use of the Hide this Page button on the MoJ claim legal aid service during lockdown.

The Design System team agreed with the component’s proposers that there should be a consistency of approach in design and functionality across Government.

Research within the context of trauma and crisis

Researching with people who have or are experiencing trauma needs careful planning. We have a duty of care to our participants and were mindful to avoid extractive research or re-traumatising those who had given their time to us. Collaborating with Hyndburn and Ribble Valley (HARV) Move On Team in Accrington was essential in enabling us to do this. We built a dummy service based on the ‘check eligibility for child maintenance’ service and tested the component in that service with people with a lived experience of domestic abuse. We repeated the same tests with people with accessibility needs.

What we tested

This is journey map of what we tested:

Findings from people with a lived experience of domestic abuse

We found that users did:

• considered the use of Exit this page in the context of reducing the risk to their personal safety
• held conflicting needs simultaneously - to know information they had inputted couldn’t be found by a perpetrator and to be able to return to the service and start from where they left off
• had a false sense of security about what happens to information they have inputted
• want to learn more about how to keep themselves safe online

Users did not:

• understand or misunderstood what happened to their information once they had inputted it
• understand or misunderstood what happened to their search history
• read large blocks of explanatory text when using the component in the prototype
• know how to return to the journey they were on.

Findings from people with accessibility needs.

We found that among test participants with accessibility needs:

• the escape key didn’t activate Exit this Page consistently for users
• content was too dense and didn’t accurately represent the participant’s experience of activating Exit this page
• the Exit this page function hindered the journey of people who use a screen reader
• visual clues (eg colour) shouldn’t be relied on to tell users how Exit this page functions
• users need instant feedback about what is happening when they activate Exit this page
• keyboard users need to be able to jump to the button quickly
• users did not consider the use of Exit this page in the context of reducing risk to personal safety
• screen reader users on iOS mobile devices navigated with a ‘screen curtain’ turned on, turning the device’s screen black and essentially hiding their screen from view

Findings common to both groups

We found that users:

• have pre-existing strategies to leave a page and/or hide information they are looking at
• don’t digest dense blocks of instructive text unless they are directed to
• couldn’t return to the journey independently
• didn’t realise that a second tab had opened when they activated Exit this page
• hide this Page, the original name for the component, didn’t accurately reflect what the component does.

Testing of the Exit this page keyboard shortcut

Using the escape key as a keyboard shortcut was found to be problematic. In some situations, other functions of the Esc key would take precedence over the Exit this Page functionality.

For example, it would first close any popovers, dialog windows, or exit out of full screen mode before it would activate EtP, requiring a user to potentially have to wait or press Esc even more times before they could actually exit the page.

Pressing the Esc key also isn't counted as being an intentional user interaction (versus clicking the mouse or pressing other keys), which made the browser behave differently depending on if the user had only pressed Esc, or had done other things before pressing Esc—which is weird and inconsistent. because of how ESCAPE is interpreted by HTML and the way it functions in some browsers.

So… SHIFT press three times was decided on as a compromise.