Psychology of Computing: Crash Course Computer Science #38

We’ve spent most of this series talking about computers. Which makes sense - this is Crash Course COMPUTER SCIENCE after all. But at their core computers are tools employed by humans and humans are pretty complicated. So today, we’re going to discuss some psychological considerations in building computers like how to make them easier for humans to use, the uncanny valley problem when humanoid robots gets more and more humanlike, and strategies to make our devices work better with us by incorporating our emotions and even altering our gaze. Oh, and we'll talk about Carrie Anne's all time favorite user interface design principle - knurling.

Want to know more about Carrie Anne?

Want to find Crash Course elsewhere on the internet?

[WillerASCruz]

This course has grown to be one of the best crash courses ever, IMHO.

[armorsmith]

Code is also an interface and if you write code, be kind to those who will read it, including your future self.

[alishaokeeffe]

I am doing a major in psychology and artificial intelligence. This was some good information

[IceMetalPunk]

Relevant anecdote: I once took a uni course on professional writing for grant proposals, and one of our projects was to basically write an actual grant proposal for whatever field we were studying. As a comp sci major, I went with the idea of using pathfinding algorithms to optimize dynamic workflow layouts in software suites with the goal of enhancing usability and productivity. (Translation: having windows and panels optimally rearrange themselves when you open and close them to maintain your preferred layout without creating overlap for you to have to deal with.) Little did I know that my writing professor had studied human-computer interaction back when she was at uni, and she loved the idea!

I didn't consider it a worthwhile goal, since I was just trying to pass the course (and honestly, I'm not so sure it was actually a good idea), so I never sent the proposal out to anyone. I do occasionally wonder if anyone would fund that research, though; I don't have more than the basic programmer's experience with HCI, though, to really know for sure.

[treymedley]

The Facebook study wasn't controversial because of what it showed. It was controversial because Facebook didn't even do a minimum of ethical approval. Participants did not consent to a study (at least not knowingly). So while it was legal (because the law doesn't care about whether you are aware of what you agree to) it was definitely not ethical. TL;DR: The FB study would have had the researcher kicked out of any accredited school anywhere for ethical violations.

[greyareaRK]

That was amazing and deserves its own series. There were a number of things I didn’t know, and I thought I was keeping up to date. Thank you!

[GrantCelley]

Please do an episode on open source software

[TheCaptain]

Love the triangle wave in the background instead of the sinusoid lol

[crazyphysicsdev]

This series is good watched every video in one day THANK YOU!!!

[rparl]

Because we're the ones in charge, FOR NOW.

[paulgap]

Speaking of a bad GUI: Facebook (in general) settings (in particular)

[TheYumbacon]

I'm a psychology and comp sci double major, and affective computing and HRI sound so cool!

[JulieHerrick]

I love the shoutout to ZORK. Oh that brings back memories of my childhood.

[halflifegordanjpog]

I have recently finished a instruction to HCI (human computer interaction) in my degree. In this video, it did mention, areas I've been tought, a good video from crash course.

[kennedya]

Finally! A comsci episode i can understand.

[XRaym]

Very nice subject ! There is a lot more in Donald Norman books about Human interactions with interface, he spoke a lot about errors prevention. For eg, imagine you have a form to fill. Rather than let the user badly fill it and sent it and then send it back on the form page if it has a missing field, prevent sending if there is empty fields. Prevent filling with erroneous data in a first place. This is way less frustrating for the user... and safer ! :P For the door example, Don explained that door which are on the street should open to the outside, so that people can escape fire inside the building faster and without having to think. Good error preventions can save lives :P Very good reading I recommend !

[remuladgryta]

Modern GUIs have generally gotten pretty crappy affordances with the rise of minimalism. Buttons no longer have rectangular beveled outlines and even regular looking text elements often offer some interaction. Often times the only way to know if an element is interactive or not is to try to interact with it.

[itsamemari]

Thank you for the video. I really appreciate it!

[woodchuck]

I thought the Facebook study was controversial because they didn't have consent to experiment on the users in the first place. Not telling your subjects that they are participating in an expirement is preaty unethical.

[Kakurady]

Don Norman would call virtual buttons having the appearance of real buttons a signifier - he now uses "affordances" for something more specific (I think that virtual buttons can be clicked, even when it doesn't look like it can, would be one example).