CPT208-Human-Centric Computing: Field Study and Analytics实地研究与分析

回顾：Recap

- 评估方法示例: Evaluation methods examples

方法 类型	受控环境 Controlled settings	自然环境 Natural settings	无需用户参与 Without users
示例	·可用性测试  Usability testing ·实验设计 Experimental design	·实地研究 Field study	·启发式评估   Heuristic evaluation  ·A/B 测试  A/B testing ·数据分析  Analytics ·预测模型  Predictive models

观察法Observations

1.观察类型：Observation types

- 直接观察 Direct observation

  ·在现场进行In the field

  ·在受控环境中进行 In controlled environments

- 间接观察（跟踪用户活动）Indirect observation: tracking users’ activities

  ·日志记录（用户日记）Diaries 

  ·交互行为记录 Interaction logging

  ·通过无人机或其他设备远程收集的视频和照片

Video and photographs collected remotely by drones or other equipment

2.直接观察

2.1实地直接观察 Direct observation in the field

- 三个易于记忆的关键要素：

·人物（The person）：是谁？

·地点（The place）：在哪里？

·事件/事物（The thing）：观察什么？

- 用于指导观察的详细框架（Robson，2014）

·空间（Space）：物理空间是什么样的？布局如何？What is the physical space like and how is it laid out?

·参与者（Actors）：有哪些人参与？他们的姓名和相关背景信息是什么？What are the names and relevant details of the people involved?

·活动（Activities）：这些参与者在做什么？为什么？What are the actors doing and why?

·物体（Objects）：现场有哪些物理对象，如家具等？What physical objects are present, such as furniture?

·行为（Acts）：有哪些具体的个人动作？What are specific individual actions?

·事件（Events）：你观察到的内容是否属于某个特别的事件？Is what you observe part of a special event?

·时间（Time）：事件的时间顺序是怎样的？What is the sequence of events?

·目标（Goals）：参与者试图完成什么目标？What are the actors trying to accomplish?

·情绪（Feelings）：整个小组和个体的情绪状态如何？What is the mood of the group and of individuals?

- 决定你的参与程度Decide on how involved you will be

·从被动观察者到积极参与者from passive observer to active participant

- 如何获得被接受 How to gain acceptance

- 如何处理敏感话题，例如文化、私人空间等How to handle sensitive topics, for example, culture, private spaces, and so on

- 如何收集数据：How to collect the data:

·收集哪些数据What data to collect

·使用什么设备What equipment to use

·何时停止观察When to stop observing

2.2人种志研究Ethnography  

- 介绍

        ·人种志是一种包含一套技术的方法论，这些技术包括参与式观察和访谈。

Ethnography is a philosophy with a set of techniques that include participant observation and interviews

        ·人种志研究者会沉浸在他们所研究的文化之中。Ethnographers immerse themselves in the culture that they study

        ·参与是关键。Participation is the key

        ·相比其他方法，它能获得更丰富、更详尽的信息，但代价高昂且具有挑战性。Richer and more detailed than other methods, but expensive and challenging

- 拓展More on Ethnography

        ·需要被观察者的配合Co-operation of people being observed is required

        ·线人非常有用 Informants are useful

        ·数据分析是持续进行的Data analysis is continuous

        ·随着理解的加深，问题会不断被完善Questions get refined as understanding grows

        ·报告通常包含具体示例 Reports usually contain examples

- 在线人种志研究Online Ethnography

        ·虚拟的、在线的、网络人种志Virtual, Online, Netnography

        ·包括线上和线下的活动Online and offline activity

        ·在线互动不同于面对面互动Interaction online differs from face-to-face

        虚拟世界具有实体世界所不具备的持续性 Virtual worlds have a persistence that physical worlds do not have

        ·伦理考量和结果呈现方式也有所不同 Ethical considerations and presentation of results are different

- 可能收集的观察内容与材料（Crabtree, 2003）Observations and materials that might be collected (Crabtree, 2003)

        ·活动或工作描述 Activity or job descriptions

        ·规范特定活动的规则和流程Rules and procedures that govern particular activities

        ·所观察活动的详细描述Descriptions of activities observed

        ·各方之间对话的录音Recordings of the talk taking place between parties

        ·参与者的非正式访谈，用以解释所观察活动的细节Informal interviews with participants explaining the detail of observed activities

        ·物理布局图，包括物品的位置 Diagrams of the physical layout, including the position of artifacts

        ·观察活动时收集的其他信息：

                >工件的照片（如文件、图表、表格、计算机等）Photographs of artifacts (documents, diagrams, forms, computers, and so forth) 

                >工件的视频Videos of artifacts

                >工件的描述Descriptions of artifacts

                >展示任务顺序的工作流程图Workflow diagrams showing the sequential order of tasks

                >展示活动之间关联的流程图Process maps showing connections between activities

2.3在受控环境中的直接观察Direct observation in a controlled environment

- 思维大声说出法Think aloud techniques

        ·这是一种让用户在操作时边做边说的方法。用户需要说出他们正在做什么、为什么这么做、他们的想法或疑惑。研究人员通过这些话语更深入地了解用户的思维过程、行为动机以及在使用系统时遇到的困难。

3.间接观察Indirect observation

- 追踪用户活动Tracking users’ activities

        ·日记记录Diaries

        ·交互日志Interaction logs

        ·网站分析Web analytics

- 在直接和间接观察中，都会使用视频、音频、照片和笔记来捕捉数据Video, audio, photos, and notes are used to capture data in both direct and indirect observations

- 示例：火车票购买  train ticket purchasing

        ·你将如何通过观察法设计一项研究，以了解用户的火车票购买体验？How will you design a study to understand users‘ train ticket purchasing experience using observations?

                >使用12306购票？

                >在火车站现场购票？

实地研究Field study

1.介绍

- 实地研究在自然环境中进行Field studies are done in natural settings

- “在野外”（in the wild）指的是原型在自然环境中被自由使用的情况“In the wild” is a term for prototypes being used freely in natural settings

- 旨在了解用户的自然行为以及技术对他们的影响 Seek to understand what users do naturally and how technology impacts them

- 实地研究在产品设计中的用途包括：Field studies are used in product design to:

        ·识别新技术的应用机会Identify opportunities for new technology

        ·确定设计需求Determine design requirements

        ·决定如何最佳地引入新技术Decide how best to introduce new technology

        ·评估技术在实际使用中的表现Evaluate technology in use

2.实地研究示例：Painpad

2.1背景

- 监测病人的疼痛程度是医生面临的一个已知难题Monitoring patients’ pain is a known challenge for physicians

- Painpad 是一个按键式设备 Painpad is a keypad device

        ·在投入两家医院使用之前，它在实验室中进行了广泛的可用性测试It was usability tested extensively in the lab before brought into two hospitals

- 研究的目标是了解 Painpad 在自然环境中、以及在英国两家医院日常工作流程中是如何被使用的Goal was to understand how Painpad was used in the natural environment and as part of routines in two UK hospitals.

2.2数据收集与参与者情况Data collection and participants

- 两项研究在两家医院进行Two studies in two hospitals

        ·共涉及54人（13名男性，41名女性）Involving 54 people (13 males, 41 females) 

        ·住院时间为1至7天，平均住院时间为2-3天Hospital stay ranged from 1-7 days, with a mean stay of 2-3 days

        ·年龄范围为32至88岁，平均年龄64.6岁，中位数64.5岁Aged between 32-88, mean and median age 64.6, 64.5

- 病人在术后被提供 Painpad，并被提示每两小时报告一次疼痛等级，护士也会同步记录评分Patients given Painpad after surgery and prompted to report pain levels every two hours, nurses also collected scores

- 其中一家医院的病人在出院时填写了用户满意度调查问卷Patients in one hospital were given a user-satisfaction survey when they left

- 同时使用1-5的李克特量表对 Painpad 进行评分Also rated Painpad on a 1-5 Likert scale

        ·隐私问题是一个重要考量Privacy was a important concern

2.3数据分析与呈现 Data analysis and presentation

- 共收集了三类数据：Three types of data were collected:

        ·对 Painpad 满意度的数据来自问卷回复Satisfaction with Painpad was based on questionnaire responses

        ·病人对每两小时一次记录的依从情况 Patients’ compliance with the two-hour routine

        ·Painpad 所收集的数据与护士收集的数据的对比How data collected from Painpad compared with data collected by nurses

- 数据显示：Data showed

        ·病人对 Painpad 的满意度在李克特量表上的评分为 4.63 Satisfaction with Painpad 4.63 on Likert scale

        ·病人的依从性表现不一：有些人喜欢使用，有些人不喜欢，或者没有注意到提示Patience compliance was mixed: some liked it while others disliked or didn’t notice the prompts

        ·病人使用 Painpad 记录的评分次数多于护士记录的次数Patients recorded more scores with Painpad than through the nurses

没有用户的直接参与without direct involvement of users

1.分类：

- 专家知识被编码为启发式规则Experts knowledge codified in heuristics

        ·启发式评估Heuristic evaluation

        ·操作演练Walkthroughs 

- 远程收集数据Data collected remotely

        ·数据分析Analytics 

        ·A/B 测试A/B testing

- 用于预测用户表现的模型Models that predict users’ performance

        ·预测建模Predictive modeling

2.分析Analytics

- 用户的各种操作可以通过软件自动记录，包括：

A variety of users’ actions can be recorded by software automatically

        ·按键记录Key presses

        ·鼠标或其他指针设备的移动Mouse or other pointing device movements

        ·浏览网页或查看帮助系统所花费的时间 Time spent searching a web page, looking at help systems

        ·在软件模块中的任务流程Task flow through software modules

        ·屏幕使用时间Screen time

        ·在虚拟环境中的移动情况 Movement in a virtual environment

- 自动记录用户活动的优点 Advantages of logging activity automatically

        ·只要不影响系统性能，它就是非侵入式的It is unobtrusive provided the system’s performance is not affected

        ·可以自动记录大量数据，之后使用可视化等工具进行深入探索与分析 Large volumes of data can be logged automatically and then explored and analyzed using visualization and other tools.

- 缺点Disadvantage 

        ·如果在用户不知情的情况下进行观察，将引发伦理问题 It raises ethical concerns about observing participants if this is done without their knowledge

- 网站分析Web Analytics

        ·一种交互日志记录形式，用于分析用户在网站上的行为 A form of interaction logging that analyses users’ activities on website

        ·设计师利用这些分析来改进他们的设计 Designers use the analysis to improve their designs

                >当设计不能满足用户需求时，用户不会再次访问网站，成为一次性用户  When designs don’t meet users’ needs, they will not return to the site, they become one-time users

        ·网站分析可以帮助设计师追踪用户在网站上的行为：Enable designers to track the activities of users on their site

                >有多少人访问网站、停留时间有多长、他们浏览了哪些页面 How many people come to the site, how long they stay, and where they go

        ·通过用户行为，网站分析为设计师提供了网站整体表现的“全貌”Offer designers the “big picture” about how their site performs based on user activity

- 示例：

        ·获取数据：CPT208 learning mall

            ·解读分析数据Make sense of the analytics

                >在这段时间内有多少人访问了该网站？How many people visited the site during this period?

                >平均访问时间为 2 分 37 秒，你认为用户在这段时间里可能会浏览哪些内容？What do you think someone might look at in 2 minutes, 37 seconds (the average time they spent on the site)?

                >跳出率指的是仅浏览你网站一页后就离开的访客所占的比例。这本书的跳出率是多少？你认为为什么跳出率是一个对任何网站都很有价值的指标？Bounce rate refers to the percentage of visitors who view just one page of your site. What is the bounce rate for this book, and why do you think this might be a useful metric to capture for any website? 

                >用户使用哪些设备访问该网站？Which devices are being used to access the site?

                >该期间访问量最大的三个语言群体是哪三个？你如何评价它们各自的跳出率？Which were the three largest language groups during the period, and what can you say about the bounce rate for each of them?

- 其他分析工具

·Moz Analytics

追踪搜索营销、社交媒体营销、品牌活动、链接和内容营销，适用于链接管理与分析：

racks search marketing, social media marketing, brand activity, links, and content marketing, and it is particularly useful for link management and analysis

网址：http://www.moz.com

·TruSocialMetrics

追踪社交媒体指标，帮助计算社交媒体营销的投资回报率：

Tracks social media metrics, and it helps calculate social media marketing return on investment

网址：www.truesocialmetrics.com

·Clicky

一款功能全面的实时分析工具，显示单个访客及其行为，有助于了解不同人群对哪些内容感兴趣：Comprehensive and real-time analytics tool that shows individual visitors and the actions they take, and it helps define what people from different demographics find interesting

网址：www.clicky.com

·KISSmetrics

提供详细分析，展示访客在购买前、中、后的行为路径：

Detailed analytics tool that displays what website visitors are doing on your website before, during, and after they buy

网址：www.kissmetrics.com

·Crazy Egg

追踪访客的点击位置，生成点击热图，适用于网站设计、可用性分析和转化率优化：

Tracks visitor clicks based on where they are specifically clicking, and it creates click heat maps useful for website design, usability, and conversion

网址：www.crazyegg.com

·ClickTale

记录访客行为，并利用元统计生成包括鼠标移动、滚动等行为的可视化热图报告：

Records website visitor actions and uses meta-statistics to create visual heat map reports on customer mouse movement, scrolling, and other visitor behaviors

网址：www.clicktale.com

3.A/B 测试A/B Testing

- 介绍：

        ·一种大规模实验（通常涉及数千名参与者或更多）A large-scale experiment (thousands of participants or more)

        ·提供了一种评估网站、移动设备上的应用程序等的方法Offers another way to evaluate a website, application of app running on a mobile device

        ·常用于评估社交媒体应用中的设计更改Often used for evaluating changes in design on social media applications

        ·比较两组用户在两个不同版本设计中的表现差异 Compares how two groups of users perform on two versions of a design

        ·如果用户不知道自己参与了测试，可能会引发伦理问题May create ethical dilemmas if users don’t know they are part of the test

- 示例：Learning Mall 登录页面

·自变量是什么？

     > 登录页面的设计版本（例如：A版 vs B版）

·因变量是什么？

     > 用户的行为表现，例如登录成功率、登录所用时间、跳出率或用户满意度

·你有假设吗？

     >例如：“新设计的登录页面（B版）将比当前页面（A版）提高用户登录成功率并减少跳出率。”

·这是组内设计还是组间设计？

     >如果每位用户仅看到一个版本，是组间设计（between-subjects design）；

     >如果每位用户都体验两个版本，是组内设计（within-subjects design）

·你将如何处理伦理问题？

     >在测试前明确告知用户他们可能会体验不同版本的页面

     >提供隐私说明和知情同意书

     >确保数据匿名化并仅用于研究目的

     >避免任何影响用户正常访问和体验的设计

4.预测模型Predictive models

- 介绍：

        ·提供了一种在不直接涉及用户的情况下评估产品或设计的方法，成本低于用户测试 Provide a way of evaluating products or designs without directly involving users, less expensive than user testing

        ·使用公式推导出各种用户表现的衡量指标 Use formulas to derive various measures of user performance

        ·其适用性仅限于任务可预测的系统，例如语音信箱系统、智能手机和专用移动设备Usefulness limited to systems with predictable tasks, for example, voicemail systems, smartphones, and dedicated mobile devices

- 示例：费茨定律Fitts’ Law (1954) 

·费茨定律预测，使用设备指向一个物体所需的时间是目标距离和目标大小的函数。

Fitts’ Law predicts that the time to point at an object using a device is a function of the distance from the target object and the object’s size.

                                                        T = k × log₂(D/S + 1.0)

        >其中：

T 表示将指针移动到目标所需的时间time to move the pointer to a target

      D 表示指针与目标之间的距离distance between the pointer and the target

      S 表示目标的大小size of the target

k 是一个常数，约为 200 毫秒/比特k is a constant of approximately 200 ms/bit

        ·距离：目标越远且越小，定位并指向它所需的时间就越长the further away and the smaller the object, the longer the time to locate it and point to it

        ·大小：目标越大，越容易、越快被点击 the bigger the target, the easier and quicker it is to reach it

        ·这一定律可以帮助设计师决定：

                >将实体或数字按钮放在哪里where to locate physical or digital buttons

                > 应该设计多大的按钮what size to make them, and

                > 按钮之间应该放得多近how close together to put them

        ·费茨定律对于评估那些定位目标时间很重要的系统非常有用，例如智能手机、手持设备和移动设备。Fitts’ Law is useful for evaluating systems for which the time to locate an object is important, for example, smartphones, handhelds, and mobile devices.

总结Summary 

- 检查可用于评估需求、模型、功能原型或系统

Inspections can be used to evaluate requirements, mockups, functional prototypes, or systems

- 用户测试和启发式评估可能揭示出不同的可用性问题

User testing and heuristic evaluation may reveal different usability problems

- 演练法是一种细致聚焦的方法，适用于评估产品的小部分内容

Walkthroughs are a fine-grained focused method for evaluating small parts of a product

- 分析法通过收集用户在网站或产品上的活动数据，了解哪些部分被使用

Analytics involves collecting data about users activity on a website or product to see which parts are used

- A/B 测试是一种大规模实验形式A/B testing is a form of large-scale experiment

费茨定律可用于预测专家在明确任务下的无误操作表现，例如评估手持设备上的按键序列或屏幕上物体的位置Fitts’ Law can be used to predict expert, error-free performance for clearly defined tasks with limited key presses, for example, to evaluate keypress sequences for handheld devices and the position of objects on a screen