Psychology of Computing: Crash Course Computer Science #38

Summary

本视频脚本探讨了计算机科学与人类行为之间的联系。强调了在设计计算机系统时，理解人类的优点和缺点是至关重要的。设计师们运用社会、认知、行为和感知心理学原理，以提高软件的可用性，即用户有效和高效实现目标的程度。视频中讨论了人类视觉系统的特性，如何影响数据展示，以及如何利用颜色和数据分组（chunking）来提高界面设计。此外，还涉及了“可供性”（affordances）的概念，以及它如何帮助用户直观地理解如何与界面互动。视频还探讨了情感智能，包括如何通过传感器和计算模型来识别和响应用户的情感状态。最后，提出了关于计算机介导通信（CMC）、增强凝视（augmented gaze）技术，以及人机交互（HRI）的见解，并对社会媒体内容的道德考量进行了讨论。

Takeaways

• 💻 计算机是工具，而人类使用这些工具，但人类的行为是复杂且不可预测的。
• 🧠 为了构建有用、易用和令人愉快的计算机系统，需要理解计算机和人类的优势与局限。
• 🎨 可用性是指一个人造物品（如软件）能够被有效和高效地用来达成目标的程度。
• 👀 人类视觉系统对颜色强度排序很敏感，但不擅长对颜色进行排序，这影响了数据展示的设计选择。
• 🧠 人类认知能力表明，当信息被分块时，我们能更有效地阅读、记忆和处理信息。
• 📊 色彩适合于展示无序的分类数据，而不是连续数据。
• 🔑 接口设计中的一个核心概念是“可供性”（affordance），它提供了事物操作的强烈线索。
• 🚪 良好的可供性设计可以让用户仅通过观察就知道如何操作，无需图片、标签或指令。
• 📚 人类的记忆在有感官线索触发时表现得更好，这就是为什么接口使用图标来代表功能。
• 🛠️ 随着用户对界面越来越熟悉，他们会建立心智模型，从而更高效地完成任务。
• ❤️ 情感智能的计算机系统可以适应其用户的情感状态，提供更富有同理心、愉悦或令人愉快的体验。
• 🤖 人机交互（HRI）研究了人类与机器人之间的互动，以及机器人如何解读人类的社会线索。

Q & A

为什么在设计计算机系统时需要理解人类的行为特性？

为了构建有用、易用且令人愉快的计算机系统，我们需要理解计算机和人类的优势与弱点。人类的行为是多变的，有时候逻辑清晰，有时候则非理性，理解这些特性有助于设计出更符合人类使用习惯的软件。

什么是可用性，它在软件设计中的重要性是什么？

可用性是指一个人造物品，比如软件，能够被用来有效且高效地达成目标的程度。它在软件设计中至关重要，因为它直接影响用户能否顺利使用软件完成任务，以及使用过程中的效率和体验。

人类视觉系统在处理颜色时有哪些优势和劣势？

人类视觉系统擅长识别和排序颜色的强度，但对于颜色的顺序排列却表现不佳。因此，在设计数据展示时，使用颜色强度来表示连续值是合适的，但用颜色来展示需要顺序排列的连续数据则可能导致设计上的灾难。

什么是“块化”（chunking），它在界面设计中的应用有哪些？

块化是指将信息分成小的、有意义的组。人类可以更有效地阅读、记忆和处理块化的信息。在界面设计中，块化被应用于下拉菜单项和菜单栏的按钮排列，使得用户更容易视觉扫描、记忆和访问。

什么是“可供性”（affordance），它在图形用户界面设计中的作用是什么？

可供性是指物品提供的操作线索。在图形用户界面设计中，可供性帮助用户仅通过外观就知道如何操作，比如按钮看起来就是可以点击的。这减少了用户对图标功能的猜测，使得计算机界面比命令行界面更易于使用。

为什么计算机介导的通信（CMC）中，人们更倾向于自我披露？

心理学研究表明，在计算机介导的通信中，人们倾向于展示更高水平的自我披露，即透露个人信息。这可能是因为CMC提供了一种相对匿名和非面对面的交流方式，使得人们感到更自在地分享个人信息。

什么是“增强注视”技术，它在视频会议中如何提升交流质量？

增强注视是一种计算机视觉和图形软件技术，可以调整头部和眼睛的图像，使得视频中的人物看起来像是直接注视着摄像头，即远程观众。这项技术可以纠正由于摄像头通常位于屏幕上方而导致的视线问题，从而改善视频会议中的交流质量。

“恐怖谷”效应是什么，它在人机交互中有何影响？

“恐怖谷”效应是指当机器人或动画人物的外观和行为接近真人但又不完全相同时，会引起人们的不安和反感。在人机交互中，这个效应提示设计者在追求机器人外观和行为的人类化时需要谨慎，以避免引起用户的不适。

为什么人们会将情感赋予非人类对象，如计算机或机器人？

人类有将情感赋予非人类对象的倾向，这被称为拟人化。计算机和机器人因其能够执行任务或移动而更容易被拟人化。这种拟人化有助于人们与这些技术建立联系，但也对设计者提出了如何使机器人行为更自然、更符合社会规范的挑战。

情感计算是什么，它在计算机系统中的作用是什么？

情感计算是一个跨学科领域，它结合了心理学、社会学和计算机科学的知识，研究计算机系统如何识别、解释、模拟和改变人的情感状态。情感感知系统可以使用传感器捕捉语音、面部表情和生物特征数据，然后利用计算模型来估计用户的情感状态，并决定如何最佳响应以达到系统的目标。

为什么说心理学和计算机科学的结合对我们的日常生活有巨大的影响？

心理学和计算机科学的结合可以帮助我们更好地理解人机交互，设计出更符合人类认知和情感需求的计算机系统。这种结合不仅能够提升用户体验，还能在教育、医疗、娱乐等多个领域提供支持，从而极大地影响和改善我们的日常生活。

在设计计算机系统时，应如何平衡效率和用户的情感体验？

在设计计算机系统时，需要考虑到用户的长期和短期目标，以及他们的情感状态。系统设计应提供多种途径来完成任务，以适应不同用户的需求和技能水平。同时，系统还应能够适应用户的情感变化，比如在用户情绪低落时提供鼓励，在用户压力大时提供帮助，以此来平衡效率和情感体验。

Outlines

00:00

😀 计算机科学与人类行为

Carrie Anne在Crash Course Computer Science的视频中介绍了计算机科学不仅仅关注计算机的硬件和算法，还关注人类行为的复杂性。她指出，为了设计出有用、易用和令人愉悦的计算机系统，必须理解计算机和人类的优势与局限。系统设计者在创建软件时会运用社会心理学、认知心理学、行为心理学和感知心理学的原理。视频中还讨论了可用性的概念，即人制造的工件（如软件）能够高效且有效地实现目标的程度。此外，还提到了人类视觉系统的特性，如对颜色强度的敏感性，以及如何利用这些特性来设计更好的计算机界面。

05:04

🤔 界面设计中的认知与情感

视频的第二部分探讨了如何通过理解人类的认知来设计界面，例如人类如何更有效地阅读、记忆和处理信息。提到了信息块化的概念，即把信息分成小的有意义的组，这与人类的短期记忆能力有关。此外，还介绍了“可供性”（affordances）的概念，即通过界面元素的外观和感觉来暗示用户如何与之交互。视频还讨论了情感智能，即计算机系统如何适应用户的情感状态，以及如何通过传感器和计算模型来识别和响应用户的情感状态。此外，还提到了Facebook在2012年进行的一项研究，该研究发现用户在看到更多正面内容时，他们的帖子也更倾向于正面，反之亦然。

10:04

🤖 人机交互与机器人技术

第三段内容讨论了人机交互（HRI）和计算机介导通信（CMC）的研究领域。提到了人们如何对机器人的行为和形态有不同的感知，以及机器人如何解读人类的社交暗示。还探讨了机器人设计中“恐怖谷”现象，即机器人与人类相似度增加时，人们的反应会有一个不自然的下降。此外，还讨论了心理学和计算机科学的结合如何影响我们的日常生活，以及这种结合带来的伦理问题，比如计算机系统是否应该对用户说谎，以及社交媒体公司是否应该为了增加用户停留时间而筛选内容。最后，提到了心理学帮助提高计算机的可访问性，以及随着机器人技术的进步，人们对机器人互动的接受度也在增加。

Keywords

💡可用性

可用性是指人造物品（如软件）能够有效和高效地实现目标的程度。视频中提到，设计良好的界面应具备高可用性，以便用户能够顺利使用而不感到挫折。例如，一个糟糕设计的界面会妨碍用户的进展，导致他们放弃使用。

💡颜色强度

颜色强度是指颜色的亮度或暗度。视频中解释了人类视觉系统擅长排列颜色强度，因此颜色强度是展示连续数据的好选择。而在排列颜色时，人类表现不佳，因此用颜色展示连续数据可能导致设计上的灾难。

💡分组

分组是指将项目放在一起组成小的、有意义的组。视频提到，人类在短期记忆中一般能处理七个左右的项目，因此分组有助于记忆和处理信息。这一概念被应用于计算机界面设计中，例如将菜单项分组以便用户更容易扫描和访问。

💡引导性

引导性是指设计元素提供的操作提示。例如，门把手是用来拉的，推板是用来推的。视频中提到，好的设计应通过引导性提示用户如何操作，而不需要额外的说明。图形用户界面广泛使用引导性，使得界面更容易使用。

💡识别与回忆

识别与回忆是心理学中的一个概念，指人类记忆在有感官提示时效果更好。视频中举例说，图标的使用就是为了让用户通过识别图标来理解其功能，而不需要回忆复杂的命令。例如，垃圾桶图标提示用户这里是删除文件的地方。

💡情感计算

情感计算是指计算系统能够识别、解释、模拟和改变人类情感的能力。视频中提到，情感计算可以使计算机更具情感智能，适应用户的情感状态，从而提供更具同理心和愉快的用户体验。

💡计算机中介通信

计算机中介通信指通过计算机进行的同步或异步通信。视频中提到，人们在计算机中介通信中往往更愿意透露个人信息，并且眼神接触在这些交互中非常重要。因此，研究人员开发了增强凝视技术来改善远程交流中的互动质量。

💡人机交互

人机交互研究的是人类如何与计算机和机器人进行互动。视频中提到，设计良好的机器人应能够理解和回应人类的社会线索，以避免令人尴尬的互动。例如，机器人应该遵循社会规则，如在人类面前道歉或避免踩到人类的脚。

💡诡异谷

诡异谷是指当机器人变得几乎像人类但又不完全像时，会引发人们的不适感。视频中提到，这种现象解释了为什么一些几乎逼真的机器人会让人感到怪异和不安。

💡道德考量

道德考量是指在设计和使用计算机系统时需要考虑的伦理问题。视频中提出了关于计算机是否应该撒谎、社交媒体是否应该策划内容以延长用户停留时间等问题，这些问题展示了计算机科学和心理学结合带来的复杂伦理挑战。

Highlights

计算机是人类使用的工具，需要理解人类的优缺点来设计有用、可用和愉悦的系统。

良好的系统设计师在创建软件时，采用社会、认知、行为和感知心理学的原则。

易用性是指人造物（如软件）能够有效和高效地实现目标的程度。

人类在排列颜色强度方面表现良好，但在排列颜色顺序方面表现较差，因此使用颜色显示连续数据可能是灾难性的设计选择。

分组（chunking）可以使人类更有效地阅读、记忆和处理信息，人类短期记忆通常能处理7个左右的项目。

知觉上的提示（affordances）提供了操作事物的强烈线索，如推板用于推，旋钮用于旋转。

人类记忆在有感官提示（如文字、图片或声音）时表现更好，这也是图标在界面设计中的重要性所在。

好的界面应该提供多种路径来实现目标，例如复制粘贴可以通过菜单和快捷键两种方式完成，分别适用于新手和专家。

情感计算（Affective Computing）通过识别、解释和模拟人类情感来使计算机对用户的情感状态作出适当反应。

人机交互（HCI）研究包括如眼神接触等对话技巧的影响，这些技巧可以增强参与感和实现对话目标。

人类喜欢赋予物体（如计算机）人类特征，尤其是那些会移动的物体，例如机器人。

随着机器人技术的进步，心理学帮助我们理解如何使人类更舒适地与机器人互动。

伦理问题是设计计算系统时需要考虑的重要方面，心理学可以帮助我们理解设计选择的效果和影响。

理解设计背后的心理学可以提高可访问性，使更多人能够理解和使用计算机。

心理学研究还展示了诸如眼神接触对说服、教学和引起注意力的重要性。

Transcripts

Hi, I’m Carrie Anne, and welcome to Crash Course Computer Science!

So, over the course of this series, we’ve focused almost exclusively on computers – the

circuits and algorithms that make them tick.

Because...this is Crash Course Computer Science.

But ultimately, computers are tools employed by people.

And humans are… well… messy.

We haven’t been designed by human engineers from the ground up with known performance

specifications.

We can be logical one moment and irrational the next.

Have you ever gotten angry at your navigation system? Surfed wikipedia aimlessly?

Begged your internet browser to load faster?

Nicknamed your roomba?

These behaviors are quintessentially human!

To build computer systems that are useful, usable and enjoyable, we need to understand

the strengths and weaknesses of both computers and humans.

And for this reason, when good system designers are creating software, they employ social,

cognitive, behavioral, and perceptual psychology principles.

INTRO

No doubt you’ve encountered a physical or computer interface that was frustrating to

use, impeding your progress.

Maybe it was so badly designed that you couldn’t figure it out and just gave up.

That interface had poor usability.

Usability is the degree to which a human-made artifact – like software – can be used

to achieve an objective effectively and efficiently.

To facilitate human work, we need to understand humans - from how they see and think, to how

they react and interact.

For instance, the human visual system has been well studied by Psychologists.

Like, we know that people are good at ordering intensities of colors.

Here are three.

Can you arrange these from lightest to darkest?

You probably don’t have to think too much about it.

Because of this innate ability, color intensity is a great choice for displaying data with

continuous values.

On the other hand, humans are terrible at ordering colors.

Here’s another example for you to put in order… is orange before blue, or after blue?

Where does green go?

You might be thinking we could order this by wavelength of light, like a rainbow, but

that’s a lot more to think about.

Most people are going to be much slower and error-prone at ordering.

Because of this innate ineptitude of your visual system, displaying continuous data

using colors can be a disastrous design choice.

You’ll find yourself constantly referring back to a color legend to compare items.

However, colors are perfect for when the data is discrete with no ordering, like categorical

data.

This might seem obvious, but you’d be amazed at how many interfaces get basic things like

this wrong.

Beyond visual perception, understanding human cognition helps us design interfaces that

align with how the mind works.

Like, humans can read, remember and process information more effectively when it’s chunked

– that is, when items are put together into small, meaningful groups.

Humans can generally juggle seven items, plus-or-minus two, in short-term memory.

To be conservative, we typically see groupings of five or less.

That’s why telephone numbers are broken into chunks, like 317, 555, 3897.

Instead of being ten individual digits that we’d likely forget, it’s three chunks,

which we can handle better.

From a computer's standpoint, this needlessly takes more time and space, so it’s less

efficient.

But, it’s way more efficient for us humans – a tradeoff we almost always make in our

favor, since we’re the ones running the show...for now.

Chunking has been applied to computer interfaces for things like drop-down menu items and menu

bars with buttons.

It’d be more efficient for computers to just pack all those together, edge to edge

– it’s wasted memory and screen real estate.

But designing interfaces in this way makes them much easier to visually scan, remember

and access.

Another central concept used in interface design is affordances.

According to Don Norman, who popularized the term in computing, “affordances provide

strong clues to the operations of things.

Plates are for pushing.

Knobs are for turning.

Slots are for inserting things into.

[...] When affordances are taken advantage of, the user knows what to do just by looking:

no picture, label, or instruction needed.”

If you’ve ever tried to pull a door handle, only to realize that you have to push it open,

you’ve discovered a broken affordance.

On the other hand, a door plate is a better design because it only gives you the option

to push.

Doors are pretty straightforward – if you need to put written instructions on them,

you should probably go back to the drawing board.

Affordances are used extensively in graphical user interfaces, which we discussed in episode

26.

It’s one of the reasons why computers became so much easier to use than with command lines.

You don’t have to guess what things on-screen are clickable, because they look like buttons.

They pop out, just waiting for you to press them!

One of my favorite affordances, which suggests to users that an on-screen element is draggable,

is knurling – that texture added to objects to improve grip and show you where to best

grab them.

This idea and pattern was borrowed from real world physical tools.

Related to the concept of affordances is the psychology of recognition vs recall.

You know this effect well from tests – it’s why multiple choice questions are easier than

fill-in-the-blank ones.

In general, human memory is much better when it’s triggered by a sensory cue, like a

word, picture or sound.

That’s why interfaces use icons – pictorial representations of functions – like a trash

can for where files go to be deleted.

We don’t have to recall what that icon does, we just have to recognise the icon.

This was also a huge improvement over command line interfaces, where you had to rely on

your memory for what commands to use.

Do I have to type “delete”, or “remove”, or... “trash”, or… shoot, it could be anything!

It’s actually “rm” in linux, but anyway, making everything easy to discover and learn

sometimes means slow to access, which conflicts with another psychology concept: expertise.

As you gain experience with interfaces, you get faster, building mental models of how

to do things efficiently.

So, good interfaces should offer multiple paths to accomplish goals.

A great example of this is copy and paste, which can be found in the edit dropdown menu

of word processors, and is also triggered with keyboard shortcuts.

One approach caters to novices, while the other caters to experts, slowing down neither.

So, you can have your cake and eat it too!

In addition to making humans more efficient, we’d also like computers to be emotionally

intelligent – adapting their behavior to respond appropriately to their users’ emotional

state – also called affect.

That could make experiences more empathetic, enjoyable, or even delightful.

This vision was articulated by Rosalind Picard in her 1995 paper on Affective Computing,

which kickstarted an interdisciplinary field combining aspects of psychology, social and

computer sciences.

It spurred work on computing systems that could recognize, interpret, simulate and alter

human affect.

This was a huge deal, because we know emotion influences cognition and perception in everyday

tasks like learning, communication, and decision making.

Affect-aware systems use sensors, sometimes worn, that capture things like speech and

video of the face, as well as biometrics, like sweatiness and heart rate.

This multimodal sensor data is used in conjunction with computational models that represent how

people develop and express affective states, like happiness and frustration, and social

states, like friendship and trust.

These models estimate the likelihood of a user being in a particular state, and figure

out how to best respond to that state, in order to achieve the goals of the system.

This might be to calm the user down, build trust, or help them get their homework done.

A study, looking at user affect, was conducted by Facebook in 2012.

For one week, data scientists altered the content on hundreds of thousands of users’

feeds.

Some people were shown more items with positive content, while others were presented with

more negative content.

The researchers analyzed people's posts during that week, and found that users who were shown

more positive content, tended to also post more positive content.

On the other hand, users who saw more negative content, tended to have more negative posts.

Clearly, what Facebook and other services show you can absolutely have an affect on

you.

As gatekeepers of content, that’s a huge opportunity and responsibility.

Which is why this study ended up being pretty controversial.

Also, it raises some interesting questions about how computer programs should respond

to human communication.

If the user is being negative, maybe the computer shouldn’t be annoying by responding in a

cheery, upbeat manner.

Or, maybe the computer should attempt to evoke a positive response, even if it’s a bit

awkward.

The “correct” behavior is very much an open research question.

Speaking of Facebook, it’s a great example of computer-mediated communication, or CMC,

another large field of research.

This includes synchronous communication – like video calls, where all participants are online

simultaneously – as well as asynchronous communication – like tweets, emails, and

text messages, where people respond whenever they can or want.

Researchers study things like the use of emoticons, rules such as turn-taking, and language used

in different communication channels.

One interesting finding is that people exhibit higher levels of self-disclosure – that

is, reveal personal information – in computer-mediated conversations, as opposed to face-to-face

interactions.

So if you want to build a system that knows how many hours a user truly spent watching

The Great British Bakeoff, it might be better to build a chatbot than a virtual agent with

a face.

Psychology research has also demonstrated that eye gaze is extremely important in persuading,

teaching and getting people's attention.

Looking at others while talking is called mutual gaze.

This has been shown to boost engagement and help achieve the goals of a conversation,

whether that’s learning, making a friend, or closing a business deal.

In settings like a videotaped lecture, the instructor rarely, if ever, looks into the

camera, and instead generally looks at the students who are physically present.

That’s ok for them, but it means people who watch the lectures online have reduced

engagement.

In response, researchers have developed computer vision and graphics software that can warp

the head and eyes, making it appear as though the instructor is looking into the camera

– right at the remote viewer.

This technique is called augmented gaze.

Similar techniques have also been applied to video conference calls, to correct for

the placement of webcams, which are almost always located above screens.

Since you’re typically looking at the video of your conversation partner, rather than

directly into the webcam, you’ll always appear to them as though you’re looking

downwards – breaking mutual gaze – which can create all kinds of unfortunate social

side effects, like a power imbalance.

Fortunately, this can be corrected digitally, and appear to participants as though you’re

lovingly gazing into their eyes.

Humans also love anthropomorphizing objects, and computers are no exception, especially

if they move, like our Robots from last episode.

Beyond industrial uses that prevailed over the last century, robots are used increasingly

in medical, education, and entertainment settings, where they frequently interact with humans.

Human-Robot Interaction – or HRI – is a field dedicated to studying these interactions,

like how people perceive different robots behaviors and forms, or how robots can interpret

human social cues to blend in and not be super awkward.

As we discussed last episode, there’s an ongoing quest to make robots as human-like

in their appearance and interactions as possible.

When engineers first made robots in the 1940s and 50s, they didn’t look very human at all.

They were almost exclusively industrial machines with no human-likeness.

Over time, engineers got better and better at making human-like robots – they gained

heads and walked around on two legs, but… they couldn’t exactly go to restaurants

and masquerade as humans.

As people pushed closer and closer to human likeness, replacing cameras with artificial

eyeballs, and covering metal chassis with synthetic flesh, things started to get a bit...

uncanny... eliciting an eerie and unsettling feeling.

This dip in realism between almost-human and actually-human became known as the uncanny valley.

There’s debate over whether robots should act like humans too.

Lots of evidence already suggests that even if robots don’t act like us, people will

treat them as though they know our social conventions.

And when they violate these rules – such as not apologizing if they cut in front of

you or roll over your foot – people get really mad!

Without a doubt, psychology and computer science are a potent combination, and have tremendous

potential to affect our everyday lives.

Which leaves us with a lot of question like you might lie to your laptop, but should your

laptop lie to you?

What if it makes you more efficient or happy?

Or should social media companies curate the content they show you to make you stay on

their site longer to make you buy more products?

They do by the way.

These types of ethical considerations aren’t easy to answer, but psychology can at least

help us understand the effects and implications of design choices in our computing systems.

But, on the positive side, understanding the psychology behind design might lead to increased

accessibility.

A greater number of people can understand and use computers now that they're more intuitive

than ever.

Conference calls and virtual classrooms are becoming more agreeable experiences.

As robot technology continues to improve, the population will grow more comfortable

in those interactions.

Plus, thanks to psychology, we can all bond over our love of knurling.

I’ll see you next week.