↪ Replying to @TheNewThinkerr
@TheNewThinkerr LucidChart lucidchart.com
Tagged: Twitter
↪ Replying to @unknown
@gerrygaffney Wow, the future is here. But I can't read all of it 😬
Schwarzman is CEO of Blackstone, a company criticized for human rights violations and rainforest deforestation. Also the company that Swedish sustainability champions(?) Oatly chose for funding. Schwarzman appears to not understand vote counting.
businessinsider.com/blackstones-sc…
All of this is important reading. x.com/venikunche/sta…
Really well-written walk-through of what's happening in Peru right now. x.com/triciawang/sta…
↪ Replying to @lyreskog
@lyreskog Agree. Though I'm consciously trying to avoid that phrase in search of something more specific. This work by @nerdydivadesign was a wake-up call for me regarding how leisurely I use some terminology.
x.com/axbom/status/1…
Crying like a baby as Ewan and @charleyboorman drive into Los Angeles on the last leg of #LongWayUp. Because that's who I am.
@longwaytv
Sneaky. The only way to only accept the "Strictly Necessary" cookies is to select one of the other options and deselect it again.
Note how the button first says "Accept All" and then switches to "Save & Close".
#covertnudging
↪ Replying to @wickedgeekie
@wickedgeekie Blog about it.
@wickedgeekie Thinking more about it, would it be possible to treat it as an actual case, where you reach exactly those aforementioned conclusions and offer a better way of posing the problem?
Tired of random LinkedIn invites? You can set LinkedIn to only allow invites from people who know your e-mail. And then, as I have, you can also set your e-mail to an address nobody knows. This means people will actually need to talk to you first.
#techsanity
↪ Replying to @Tynell
@Tynell The logical fallacy of course being the idea that there was ever a problem to begin with ;)
↪ Replying to @unknown
@pedrobjorn Every day is a lesson in inequality.
I can't begin to tell you how much time I save by having decided that wearing mismatched socks doesn't bother me.
I also save on socks because singles are never discarded.
↪ Replying to @JesperBylund
@JesperBylund Not necessarily. The minority in the cupcake example would really be the people with allergies. We don't test for allergies, we test for a winner of the options we put forward. Rather than declare a winner the best solution may be more choice.
x.com/axbom/status/1…
↪ Replying to @JesperBylund
@JesperBylund I do think the minority aspect is a feature of A/B-testing that should be called out. A/B-testing can (sometimes) be effective in boosting profits, but can be at the expense of select people being sidestepped - which the test will rarely show. So - yes - skewed, with real impact.
@JesperBylund And of course I'm also critical of the dogma. :)
@JesperBylund If you test blue and pink cupcakes against each other you may find more blue cupcakes eaten at the time, place and by the people you test them with. But if you don't talk to people you won't find out about allergies. And why a number of people still preferred the pink ones.
↪ Replying to @JesperBylund
@JesperBylund Not at all. I would argue it all depends on what you're trying to achieve, and what you believe the data is saying. Understanding the data is key, not a focus on declaring "winning" designs.
@JesperBylund You do need to complement statistical testing with focused qualitative research with a diversity in health, experience, status, ability, etceteras – if you are trying to achieve inclusive design.
This x 1000 🔥🔥🔥
A/B-testing excludes minorities by design.
A/B-testing promotes the way of the majority. If you are harmed by the design you may ignore it, make mistakes and/or have autonomy/health/privacy stolen from you. But it won't make a blip in an A/B-test. x.com/beantin/status…
And of course an A/B test also won't uncover harm to the majority as long as it's hidden from them. How well that harm is hidden is perfectly optimized with an—you guessed it—A/B test.
↪ Replying to @yungtsai
@yungtsai @beantin @per_p @Windows Yeah it"s not perfect. I'd rather redesign the whole thing. My thinking was rhat Confirm is assertive (I"m confirming my selection of 'no'), whereas Accept is submissive (I"m accepting someone else's suggestion). But yes, even more clarity in this interface is needed.