- From: <josh@interaccess.ie>
- Date: Mon, 11 Jul 2016 10:15:36 +0000
- To: w3c-wai-gl@w3.org, "team-wcag-editors@w3.org" <team-wcag-editors@w3.org>
- Message-Id: <emb35bf4a9-6d52-47e2-858a-f37d7abf2bdd@josh_machine>
[Chair hat off] In a previous thread the issue of user testing with people with cognitive impairments has been brought up. I thought this could be a good time to share some of my own thoughts in this area. As some of you may know, I ran a user testing lab in the National Council for the Blind of Ireland for around 10 years. During that time I ran user tests with a wide range of users, not just those who are blind/VIP. In that time I did a small amount of testing with people with cognitive impairments, mental health issues etc. I found this a difficult group to test with due to my own concern about the ethics of doing so properly. One very strong reservation I have about this whole area is simple. Does the user have the ability the objectively separate the tasks they are asked to perform in a test (and the natural success/failure when trying to completing these tasks) from their overall 'sense of self'? What I mean is that will the user be able to realise that their actions are being objectively observed without any 'judgement' on their performance? I would hate to think that a user would come away from a user test, where many tasks were failed (which is great usability information) but feeling worse about themselves, or as if _they_ were some kind of failure. In short, I think user testing is a bit of a performance, within an utterly contrived environment. Some people take to this well, others don't. This is a very thorny issue but one I want to flag. I don't think testing should take place at all without a strong framework about how to deal with these sensitive situations. Sometimes you may have to make a call not to test, if it isn't in the best interest of the user test participant, even if they may be a perfect candidate for 'rich data'. I've made this call not to test in the past, and my overall take away was that I am largely uncomfortable doing this kind of testing, unless I'm sure it is in the participants best interest, separately from whatever the secondary need of a project/client is. Thoughts? Josh
Received on Monday, 11 July 2016 10:13:29 UTC