红色药丸 蓝色药丸

近日，社交媒体上一道“生死抉择题”引发热议。

无注释原文：

Writing A Strong Survey: Quit Asking Leading Questions

From: Forbes

Dec 2, 2022

I own a market research company, so I understand the importance gathering data from your target audience and using that data in order to make product development, marketing and strategic business decisions. I also know that surveys can provide quantitative, representative information that's extremely valuable to our clients.

But the quality of the data you get as an output from surveys is only as strong as how well you craft the survey questions. One of the main pitfalls I see teams make is writing leading questions that might get the company results they want to hear, but those results are not always accurate.

What's a 'leading question?'

In short, a "leading question" is a question that subtly (or not-so-subtly) leads a participant to a response the company wants to hear. Unless you are purposefully trying to use your data to make a point (and I've seen plenty of these types of surveys in an election year), your goal should be to leave out any questions that would lead the survey participant to respond in a certain way.

Below are common ways I've seen companies ask leading questions and how to avoid them:

1. Don't make assumptions for your participants.

“How well do you think we delivered on your project?” implies your team did a good job. To make it more objective, rephrase to ask, “How did we deliver on your project?” It's a subtle change, but the second question takes out the implicit bias.

2. Balance the options from which participants can choose.

Let's say you want to know the level of satisfaction or dissatisfaction customers have with your product or service. If you ask this in a multiple-choice format, you need to ensure the scale is balanced for both positive and negative responses. The following options, for example, would not work:

• Extremely satisfied

• Very satisfied

• Satisfied

• Somewhat unsatisfied

• Unsatisfied

As a rule of thumb, if you have “extremely” on the positive end, you should have “extremely” on the negative side as well, and give a neutral option in the middle of the scale.

3. Don't 'lead the witness' (or participant).

Again, if you're seeking truth (and you should have started with knowing why you're conducting the survey in the first place), you don't want to put statements in front of a question that will lead participants to respond in a certain way.

For instance, my company works a lot with educational technology companies. We would never write a question such as this: “Most of the people who use this educational technology product say that the adaptive and personalized nature of the product is better than others on the market. What do you think?” This is because, now, the participant knows exactly what we want them to think.

4. Avoid double-barrel questions.

Be mindful about including two questions in one and limiting how a participant can respond. Say you ran an event for customers; you wouldn't want to ask, “What do you think about the location and the quality of the speakers?” and then have multiple choice options where they are really only able to respond to one question.

They might have loved the venue but hated your presenters. The integrity of the data you gather is compromised when you put questions together.

5. Don't write coercive questions.

These questions don't sit well. In my experience, they tend to come up more in customer satisfaction surveys and employee surveys when you really want to be able to highlight the positives. Examples of coercive questions include: "You'll leave us a positive Google Review, won't you?" "You'll tell your friends to visit our restaurant, right?" Naturally, these are full of bias, so they should be avoided at all costs.

Know what the goals of your survey are. If you want to get data that is not objective and that includes biases, yes, leading questions can help you get those data points. But the integrity of the data cannot be trusted when you're not writing questions in an objective way for your participants to answer.

- ◆ -

注：完整题目见本文开头；中文文本为机器翻译并非一一对应，仅供参考

含注释全文：

Writing A Strong Survey: Quit Asking Leading Questions

From: Forbes

Dec 2, 2022

I own a market research company, so I understand the importance gathering data from your target audience and using that data in order to make product development, marketing and strategic business decisions. I also know that surveys can provide quantitative, representative information that's extremely valuable to our clients.

我经营了一家市场调研公司，因此我深知从目标受众那里收集数据并利用这些数据进行产品开发、营销和战略业务决策的重要性。我也知道，调查能够提供定量的、有代表性的信息，这些信息对我们的客户非常有价值。

quantitative /ˈkwɒn.tɪ.tə.tɪv/ 表示“定量的”，英文解释为“relating to numbers or amounts”如：quantitative analysis 定量分析。

But the quality of the data you get as an output from surveys is only as strong as how well you craft the survey questions. One of the main pitfalls I see teams make is writing leading questions that might get the company results they want to hear, but those results are not always accurate.

但从调查中获得的数据质量，很大程度上取决于调查问题的设计。我经常发现团队所犯的主要错误就是提出引导性问题，这些问题可能会得到公司想要听到的结果，但这些结果并不总是准确的。

结合我们常见的aircraft/spacecraft/hovercraft等词可以猜到，craft可以表示“船，艇；飞行器，航空器”，英文解释为“a vehicle for travelling on water or through the air”，如：naval craft 军舰，civilian craft 民用船，cpatrol craft 巡逻艇，rescue craft 救生艇。

📍补充：hovercraft 气垫船，military aircraft 军用飞机，a manned/unmanned spacecraft 载人/无人的宇宙飞船。

📍此外，craft作名词还可以指“工艺，手艺；技术”，英文解释为“skill and experience, especially in relation to making objects; a job or activity that needs skill and experience, or something produced using skill and experience”，如：the craft of furniture making/sewing 家具制作/缝纫工艺 political/literary craft 政治手腕／文学手法，rural/ancient/traditional crafts 乡村的/古老的/传统的手艺，craft workers (= skilled workers) 熟练工人，工匠。

📍当然，craft也可以作为动词表示“精制，（手工）制作”，举个🌰：Each doll is crafted individually by specialists. 每个布娃娃都由专业人员手工制作。

🎬电影《外星人保罗》（Paul）中的台词提到：Hey, imagine if we got buzzed by a craft. 想象一下如果一艘飞船从我们头上掠过。

表示“隐患；陷阱”，英文解释为“a likely mistake or problem in a situation”举个🌰：There's a video that tells new students about pitfalls to avoid. 有一个视频讲述了新生应避免的错误。

leading question /ˌliː.dɪŋ ˈkwes.tʃən/ 表示“诱导性的问题”，英文解释为“a question that tricks someone into answering in a particular way”

What's a 'leading question?' 什么是“引导性问题”?

In short, a "leading question" is a question that subtly (or not-so-subtly) leads a participant to a response the company wants to hear. Unless you are purposefully trying to use your data to make a point (and I've seen plenty of these types of surveys in an election year), your goal should be to leave out any questions that would lead the survey participant to respond in a certain way.

简而言之，“引导性问题”是一种隐晦地（或直白地）引导参与者给出公司想要听到的答案的问题。除非你是有意用数据来表明观点（我在选举年见过很多这类调查），你的目标应该是避免提出任何会引导参与者以特定方式回答的问题。

subtly /ˈsʌt.əl.i/ 表示“微妙地，精妙地;巧妙地”，英文解释为“in a way that is small and difficult to notice or describe, but important”举个🌰：His expression changed subtly. 他的表情微妙地变了。

Below are common ways I've seen companies ask leading questions and how to avoid them:

以下是我看到的公司提出引导性问题的常见方式，以及如何避免它们：

1. Don't make assumptions for your participants. 不要为参与者做出假设。

“How well do you think we delivered on your project?” implies your team did a good job. To make it more objective, rephrase to ask, “How did we deliver on your project?” It's a subtle change, but the second question takes out the implicit bias.

“您觉得我们在您的项目上做得有多好？”暗示了你们的团队做得很好。为了使其更加客观，可以换一种问法：“我们在您的项目上的表现如何？”这是一个细微的变化，但后一个问题避开了隐含的偏向性。

deliver /dɪˈlɪv.ər/ 1）表示“运送，递送，投递”，英文解释为“to take goods, letters, parcels, etc. to people's houses or places of work”举个🌰：Mail is delivered to our office twice a day. 邮差每天给我们办公室投递两次邮件。

2）表示“给…助产；接生；分娩”，英文解释为“to (help) give birth to a baby”举个🌰：The baby was delivered by a midwife. 婴儿在助产士的帮助下出生了。

3）表示“实现；产生；兑现（诺言）”，英文解释为“to achieve or produce something that has been promised”举个🌰：It has failed to deliver (what it promised). （承诺）没有兑现。

作名词，objective /əbˈdʒek.tɪv/ 表示“目标；目的”，英文解释为“something that you plan to do or achieve”举个🌰：Can the sales team achieve/meet its financial objectives? 销售人员能否完成财务目标？

作形容词，表示“客观的；就事论事的；不带个人感情的”，英文解释为“not influenced by personal feelings or opinions; considering only facts”如：an objective analysis /assessment 客观的分析/评价。

rephrase /ˌriːˈfreɪz/ 表示“给…重新措词；改用别的措词表述；用更清楚的措词表达”，英文解释为“to say or write something again in a different and usually clearer way”举个🌰：Could you rephrase your question, please? 你能否把你的问题更清楚地表述一遍？

implicit /ɪmˈplɪs.ɪt/ 1）表示“不明言的，含蓄的”，英文解释为“suggested but not communicated directly”举个🌰：He interpreted her comments as an implicit criticism. 他将她的评论理解为含蓄批评。

2）表示“无疑问的；无保留的”，英文解释为“complete and without any doubts”如：implicit trust 绝对的信任。

表示“偏见；偏心；偏向”，英文解释为“a strong feeling in favour of or against one group of people, or one side in an argument, often not based on fair judgement”举个🌰：Employers must consider all candidates impartially and without bias. 雇主必须公平而毫无成见地考虑所有求职者。

2. Balance the options from which participants can choose. 平衡参与者的可选项。

Let's say you want to know the level of satisfaction or dissatisfaction customers have with your product or service. If you ask this in a multiple-choice format, you need to ensure the scale is balanced for both positive and negative responses. The following options, for example, would not work:

假设你想知道客户对你的产品或服务的满意度或不满意度。如果你以多项选择题的形式提问，你需要确保正面和负面选项都是平衡的。例如，以下选项是不合适的：

• Extremely satisfied 极其满意

• Very satisfied 非常满意

• Satisfied 满意

• Somewhat unsatisfied 有点不满意

• Unsatisfied 不满意

1）表示“标度，刻度；等级”，英文解释为“a set of numbers, amounts, etc., used to measure or compare the level of something”如：the Centigrade/Fahrenheit scale 摄氏/华氏温标。

2）表示“比例；比例尺”，英文解释为“the relation between the real size of something and its size on a map, model, or diagram”如：a scale of 1:10,000 比例尺为1:10000。

3）表示“大小；规模；范围”，英文解释为“the size or level of something, especially when this is large”举个🌰：We don't yet know the scale of the problem. 我们还不知道问题有多严重。

As a rule of thumb, if you have “extremely” on the positive end, you should have “extremely” on the negative side as well, and give a neutral option in the middle of the scale.

按照经验，如果在正面反馈中有“极其”这个词，负面反馈中也应该有这个词，并在所有选项中间给出一个中性选项。

表示“实用的估测方法；经验之谈；经验法则”，英文解释为“a practical and approximate way of doing or measuring something”举个🌰：A good rule of thumb is that a portion of rice is two handfuls. 有个方便实用的估算方法，即两把生米便是够一人吃的量。

neutral /ˈnuː.trəl/ 作形容词，表示“中立的，不偏不倚的”，英文解释为“not saying or doing anything that would encourage or help any of the groups involved in an argument or war”举个🌰：It's important that I remain neutral. 我保持中立是很重要的。

作名词，表示“中立者”，英文解释为“a neutral person or thing”举个🌰：Sweden and Switzerland were neutrals during the war. 在战争期间，瑞典和瑞士是中立国。

3. Don't 'lead the witness' (or participant). 不要“诱导证人”（或参与者）。

Again, if you're seeking truth (and you should have started with knowing why you're conducting the survey in the first place), you don't want to put statements in front of a question that will lead participants to respond in a certain way.

再次强调，如果你寻求真相（首先，你应该知道你为什么要进行这次调查），你就不要在问题前面加上会引导参与者以某种方式回答的语句。

For instance, my company works a lot with educational technology companies. We would never write a question such as this: “Most of the people who use this educational technology product say that the adaptive and personalized nature of the product is better than others on the market. What do you think?” This is because, now, the participant knows exactly what we want them to think.

例如，我的公司与教育技术公司合作甚多。我们绝不会这样提问：“大多数使用这款教育技术产品的人都说，该产品的适配性和个性化特点比市场上的其他产品好。您怎么看？”因为，现在，参与者已经知道我们希望他们怎么想了。

adaptive /əˈdæp.tɪv/ 表示“适应的;有适应性的”，英文解释为“having an ability to change to suit different conditions”

表示“标明某人姓名的；为某人特制的；个性化的；个人化的”，英文解释为“used to describe an object that has someone's name on it, or that has been made for a particular person”举个🌰：His car has a personalized number plate - LR. 他的车上挂了一块标有他的姓名首字母的车牌——LR。

4. Avoid double-barrel questions. 避免使用双重问题。

Be mindful about including two questions in one and limiting how a participant can respond. Say you ran an event for customers; you wouldn't want to ask, “What do you think about the location and the quality of the speakers?” and then have multiple choice options where they are really only able to respond to one question.

注意不要将两个问题合二为一，限制参与者的回答方式。假设你为客户组织了一个活动；你不能这么问：“您认为活动地点和演讲者的质量如何？”然后给出多选选项，他们实际上只能回答其中一个问题。

double-barrel在调查或问卷的语境中指的是一个问题中包含了两个或多个子问题，但只允许回答者给出一个答案。这样的问题设计可能会导致回答者感到困惑，因为他们可能对子问题中的某一部分有不同的看法。

📍double-barrelled 表示“双管的”，英文解释为“A double-barrelled gun has two barrels (= parts shaped like tubes).”如：a double-barrelled shotgun 双管猎枪

📍barrel /ˈbær.əl/ 表示“枪管”，英文解释为“the long part of a gun that is shaped like a tube”或者表示“桶”，英文解释为“a large container, made of wood, metal, or plastic, with a flat top and bottom and curved sides that make it fatter in the middle”举个🌰：They drank a whole barrel of beer (= the contents of a barrel) at the party. 聚会上他们喝了整整一桶啤酒。

They might have loved the venue but hated your presenters. The integrity of the data you gather is compromised when you put questions together.

他们可能很喜欢场地，但不喜欢你的演讲者。当你把问题放在一起时，收集到的数据完整性就会受到影响。

1）表示“完整”，英文解释为“the quality of being whole and complete”举个🌰：A modern extension on the old building would ruin its architectural integrity.  在古老建筑物上附加扩建的现代建筑会破坏建筑风格的完整统一。

2）表示“正直；诚实；职业操守”，英文解释为“the quality of being honest and having strong moral principles that you refuse to change”举个🌰：No one doubted that the president was a man of the highest integrity. 没有人怀疑总裁是个光明磊落的人。

compromise /ˈkɒm.prə.maɪz/ 1）表示“妥协；折中；让步；和解”，英文解释为“an agreement in an argument in which the people involved reduce their demands or change their opinion in order to agree”举个🌰：Let's compromise at/on ¥10. 我们就来个折中价10元。

2）表示“损害，伤害，危及”，英文解释为“to risk having a harmful effect on something”举个🌰：We would never compromise the safety of our passengers. 我们永远不会做有损乘客安全的事。

3）表示“放弃；背离，违背；降低”，英文解释为“to allow your principles to be less strong or your standards or morals to be lower”举个🌰：Don't compromise your beliefs/principles for the sake of being accepted. 不要为了得到别人的认同就放弃了你自己的信仰/原则。

区分：

📍comprise表示“包含；由…组成”，英文解释为“If you say that something comprises or is comprised of a number of things or people, you mean it has them as its parts or members.”举个🌰：The class is comprised mainly of Chinese and Japanese students. 这个班主要由中国和日本学生组成。

🎬电影《拉瑞·克劳》（Larry Crowne）中的台词提到：They comprise 20% of the world's fresh surface water supply. 它们提供了地球表面百分之二十的淡水。

5. Don't write coercive questions. 不要提出强迫性问题。

These questions don't sit well. In my experience, they tend to come up more in customer satisfaction surveys and employee surveys when you really want to be able to highlight the positives. Examples of coercive questions include: "You'll leave us a positive Google Review, won't you?" "You'll tell your friends to visit our restaurant, right?" Naturally, these are full of bias, so they should be avoided at all costs.

这类问题让人感到不适。依我的经验，当你真的想强调积极的一面时，它们往往会出现在客户满意度调查和员工调查中。强迫性问题的例子包括：“您会给我们留下好评，对吗？”、“您会告诉您的朋友来我们的餐馆吃饭，对吗？”很明显，这些问题充满了偏向性，因此应当尽量避免。

coercive /ko'ɝsɪv/表示“强制的；胁迫的”，英文解释为“using threats or orders to make someone do something they do not want to do”如：coercive measures to reduce absenteeism 减少旷工现象的高压措施。

表示“（某人）可接受；受（某人）欢迎”，英文解释为“to be something that you agree or are pleased with”举个🌰：Their decision/answer didn't sit well with the Board of Directors. 他们的决定/答案没有得到董事会的认可。

🎬电影《性本恶》（Inherent Vice）中的台词提到：that don't sit well with me. 我不能接受这样，

Know what the goals of your survey are. If you want to get data that is not objective and that includes biases, yes, leading questions can help you get those data points. But the integrity of the data cannot be trusted when you're not writing questions in an objective way for your participants to answer.

了解你的调查目标是什么。如果你想获得不客观且包含偏向性的数据，引导性问题确实可以帮助你得到这些数据点。但当你没有以客观的方式提问让参与者回答时，数据的完整性就不可信了。

- 词汇盘点 -

- 词汇助记 By ChatGPT -