本文只是自己復(fù)習(xí)時(shí)對ppt的總結(jié)

Introduction

• CLI：Command Language Interface
• GUI：Graphical User Interface
• NUI : Natural User Interface
• DM: Direct Manipulation
• HIP: Human Information Processing
• GOMS: Goal, Operation, Method, Selection

Human

sensors

• Physical: vision, sound, touch
• Chemical: smell, taste, pain

Processing

• Processors: brain, eyes (Distributed processors )
• Memory: long-term, short term, sensory memory
• Cognition: attention, semantic, emotion, mood, …

Metaphor

• design model 設(shè)計(jì)師腦子里的模型
• user's model 用戶頭腦里的
• system image 交互的界面

Ergonomics 人體工程學(xué): Fit the body

我們能讓椅子去適應(yīng)人，也應(yīng)該讓交互界面去適應(yīng)人，人自己調(diào)整也可以

goal

• productivity
• health and safety
• satisfaction
• quality

case: keyboard

• qwerty是為了減慢速度而設(shè)計(jì)的，左手承擔(dān)了57%的工作，而且經(jīng)常要用無名指之類的
• dvorak是為了效率而設(shè)計(jì)的
• 2003年有XPeRT（而且容易從qwerty轉(zhuǎn)過來）
• 但為什么qwerty仍是主流？因?yàn)閏ost of learning

Perception

• 別用高純度的顏色
• consider the people with color vision deficiency

visual

• Searching for Patterns 比如按行還是按列
• Aerial perspective 遠(yuǎn)處的會(huì)有暈染大面積的顏色

• 會(huì)潛意識(shí)認(rèn)為光從上方來

  
  
  image.png

• Contexts and Expectations

  
  
  image.png

Cognition

The 5E principle for human interaction

• Easy to use
• Easy to learn
• Easy to remember
• Effective
• Error-free

Attention

• Cocktail party effect
• Ear-switching
• Automaticity 這部分不太需要注意力

Recovering

• Cued recall: eg. icons

HIP Model

（3 subsystems）

Subsystems' parameters:

• memory: μ-storage capacity; σ-decay time; κ-code type(acoustic/visual/...)
• processor: τ-cycle time
• principles
  • τ varies inversely with stimulus intensity
  • Movement is not continuous
  • Chunks must be added to LTM through WM.
  • Chunks' cues and relationships with other chunks in LTM is important

Perceptual system

• Visual

  
  
  image.png

• Auditory

  
  
  image.png
  
  

  Both τ=100(50~200) ms

Motor system

τ=70（30~100）ms

• Eye Movement：τ=230（70~700）ms

Cognitive system

τ=70（25~170）ms

• working memory
  σ=7（5~226）sec
  σ[1chunk]=73（73~226）sec
  σ[3chunk]=7（5~34）sec
  μ=3（2.5~4.1）chunks
• long term memory
  σ=∞；μ=∞

SUMMARY

image.png

• τ[visual] is slower for fainter screen

HIP應(yīng)用

• 平均一個(gè)單詞5個(gè)字母，一個(gè)短語2.5個(gè)單詞
• 這個(gè)模型可以用來分析很多問題，比如morse碼的接收速率等等
  更多請參見書：The psychology of Human Computer Interaction

GUI

design principles

always remember purpose

• visibility
  • grouping
    • Gestalt Theory (完形心理學(xué))
      law of closure(自動(dòng)補(bǔ)充)/similarity(相似的放在一起)/proximity(鄰近的放一起)/symmetry/continuity/
    • white space
  • order
  • decoration
    • STUDIES SHOW THAT TEXT IN ALL CAPS SLOWS READING BY 12% SO YOU SHOULD MIX CASE
  • alignment
    • align left hand --easier to read (names: surnames should be easy to scan)
• feedback

• mapping

  
  
  bad mapping
  
  
  good mapping
• affordance：to give a clue
• consistency 內(nèi)外的一致性

evaluation

observational

2.1.2.1 Simple observation
2.1.2.2 Think-aloud 讓用戶邊做邊說出自己想的
2.1.2.3 Constructive Interaction
2.1.2.4 Videotape retrospection
2.1.2.5 Query methods

experimental

• before
  ?? Their identities will be kept confidential
  ?? It’s the system that is being tested, not the user
  ?? They can quit any time they want
  ?? Any monitoring device that is being used
  ?? Any risk that may be involved
• after
  • Tool: SPSS
  • Fit to existing model/theories where applicable
    • Power Law of Practice: task time on n^th trial: Tn = T1*n^(-a)
    • Hick’s Law: for selecting action reaction time = a + blog2(C)
      (C: # of choices)
      --Giving a user many choices simultaneously is usually faster than is organizing the same choices into hierarchical groups
    • Fitt’s Law: Time for pointing to targets=a+b*log2(A/W+1)
      (A-distance; W-size of button)
  • Word Selections
    • Subjects Participants
    • Males & Females Men & Women
    • Disabled People with disabilities
    • “he/she” if you don’t really mean “he”
    • “Statistically significant” rather than “significant”
    • “The experiment data shows no effect” rather than “there is no effect”
    • Don’t report in a way that the subjects’ could be identified

optimization