答复率

播放段落

It is commonly accepted that any survey conducted by mail, or over the internet, or by telephone will have a very low response rate .  It is not unheard of for such surveys to have less than 5% of the chosen sample actually return usable results.
::人们普遍认为,通过邮件、互联网或电话进行的任何调查的回复率都很低。 这种调查只有不到所选抽样的5%实际返回了可用结果,这是众所周知的。

播放段落

It certainly seems logical to attempt to get a response rate as high as possible, but does a low response mean that the experiment is invalid?  The answer may surprise you, look for it after the example questions below.
::试图获得尽可能高的回复率似乎是符合逻辑的,但低的回复率是否意味着实验无效?答案可能会令你吃惊,在下面的示例问题之后再寻找答案。

播放段落

Response Rates
::答复率

播放段落

Response rate is expressed as the percentage of total attempts made at data collection divided by the number of successful responses.  If you conduct a survey of 50 students, and 10 of them choose not to participate, then you would say that you had a response rate of $\begin{align*}\frac{40}{50} = .80 \end{align*}$  or 80%.
::响应率表示为在数据收集中所做的全部尝试的百分比除以成功响应的数量。 如果你对50名学生进行了调查,其中10名学生选择不参加,那么你会说答复率为4050=80或80%。

播放段落

There is no actual widely accepted formula for determining an acceptable response rate.  Although it is logical and generally accepted that a higher response rate is better, there is quite a bit of debate over the attention that is generally paid to response rate as a determining factor in the validity of a particular poll .  In fact, during a study of the importance of a high response rate, a group of researchers compared the accuracy of predictions made from data collected over 15 years by telephone (with a response rate of 60%) and from data collected by mail (20% response rate).  It turned out that the mail survey, even with the much lower response rate, was more than 3 times as accurate in predictions as the telephone survey was! $\begin{align*}^\ast \end{align*}$ (Visser, Krosnick, Marquette, and Curtin – 1996)  However, other studies have shown that the results may also go the other way, with a low response rate directly resulting in poor accuracy in the results.
::在确定可接受的答复率方面,没有实际得到广泛接受的公式。 虽然符合逻辑,也普遍接受高答复率更好,但对于一般将答复率作为特定民意调查有效性的决定因素而给予的关注却存在相当的争议。 事实上,在对高答复率重要性的研究中,一组研究人员比较了15年来通过电话收集的数据(答复率为60%)和通过邮件收集的数据(答复率为20%)所作的预测的准确性。 结果显示,邮件调查,即使答复率低得多,在预测中也比电话调查准确3倍多! *(Visser、Krosnick、Marquette和Curtin-1996年),但其他研究表明,结果也可能相反,低答复率直接导致结果的准确性差。

播放段落

What all this tells us is that it is important to keep track of response rates in order to have as much data as possible to use in determining the accuracy of your results, but that even a poor response may not mean that the data is incorrect.  What is ultimately important is to limit or offset the difference that any missing responses may have on your results.
::所有这些都告诉我们,必须跟踪答复率,以便尽可能多地利用数据来确定你的结果的准确性,但即使答复不力,也并不意味着数据不正确。 最终重要的是限制或抵消任何缺失答复对结果的差别。

播放段落

A common approach to evaluating the effects of non-responders is to return to a small sample of members who initially did not respond, and make a concerted effort to get responses from them to see if the results appear significantly different from the data collected from the main group.
::评估不答复者影响的一个共同办法是,回到最初没有答复的成员的少量抽样中,并作出一致努力,从他们那里得到答复,看看结果是否与从主要组收集的数据大不相同。

播放段落

Calculating the Response Rate 
::计算响应率

播放段落

What is the calculated response rate from a survey attempting to find the most popular dog food brand of a sample of 130 dogs, if 24 of them did not eat any of the brands at all?
::一项调查试图找到130只狗样本中最受欢迎的狗食品品牌,如果其中24只狗完全不吃任何品牌,那么调查得出的计算答复率是多少?

播放段落

Response rate is calculated as:
::答复率计算如下:

播放段落

$$\begin{align*}\frac{(number \ of \ successful \ collections)}{(total \ attempts \ at \ collection)}=\frac{106}{130}=.815 \ or \ 81.5 \%\end{align*}$$
::=106130=815或81.5%

播放段落

Determining Number of Responses Needed 
::所需答复数目

播放段落

If you are conducting a study of the effects of music on homework study, and you have determined that you will need a minimum 70% response rate in order to have enough data to accurately determine the results, how many responses will you need to receive out of the 53 students you originally study?
::如果您正在对音乐对家庭作业研究的影响进行研究,而且您已经确定,您需要至少70%的回复率,以便有足够的数据准确确定结果,那么在您最初学习的53名学生中,您需要收到多少回复?

播放段落

Multiply the response rate by the total number of attempts to find the minimum number of successful collections:
::将答复率乘以试图找到最低成功收缴次数的总数:

$$\begin{align*}.70 \times 53 =37.1\end{align*}$$

播放段落

Since the minimum stated response is 70%, and since we cannot collect .1 results, round up to 38 successful data collections.
::由于所声明的回复率最低为70%,而且由于我们无法收集1个结果,因此汇集了38个成功的数据收集工作。

播放段落

Real-World Application: School Cafeteria 
::现实世界应用:学校食堂

播放段落

While conducting a study of the most popular foods to serve in the school cafeteria, you collect the following data:
::在对学校食堂最受欢迎的食品进行研究时,您可收集以下数据:

播放段落
1. Students in the school: 425
   ::在校学生:425人
播放段落
2. Students not present in the cafeteria for questioning: 225
   ::没有在餐厅接受询问的学生:225人
播放段落
3. Students preferring pizza: 130
   ::学生更喜欢比萨饼:130
播放段落
4. Students preferring hamburgers: 60
   ::学生更喜欢汉堡包:60
播放段落
5. Students preferring pasta: 10
   ::偏爱意大利面的学生:10

播放段落

Use this data to answer the following:
::使用此数据回答 :

播放段落

1. What is the calculated response rate of the study?
::1. 研究的计算答复率是多少?

播放段落

The calculated response rate is  $\begin{align*}\frac{200}{425}=.471=47.1 \%\end{align*}$
::计算答复率为200425=471=47.1%

播放段落

2. What is the most popular food?
::2. 最受欢迎的食物是什么?

播放段落

Pizza: $\begin{align*}\frac{130}{200}=.65=65 \%\end{align*}$  of responders
::比萨:130200=.65=65%的响应者

播放段落

3. Based on the response rate, would you consider the results reliable?
::3. 根据答复率,你是否认为结果可靠?

播放段落

The response rate suggests that less than half of the students in the school submitted a vote, so the results would seem suspect.
::答复率表明,该校不到一半的学生提交了投票结果,结果似乎令人怀疑。

播放段落

4. Based on all available information, would you consider the results reliable?
::4. 根据所有现有资料,你是否认为结果可靠?

播放段落

The 47.1% response rate actually represents 100% of the students who eat in the cafeteria, since the students not available for questioning went elsewhere for lunch. Considering this, the results were likely quite reliable.
::47.1%的回复率实际上代表了在餐厅吃饭的学生的100%,因为无法接受询问的学生去别处吃午餐。 有鉴于此,结果很可能相当可靠。

播放段落

Earlier Problem Revisited
::重审先前的问题

播放段落

It is commonly accepted that any survey conducted by mail, or over the internet, or by telephone will have a very low response rate.  It is not unheard of for such surveys to have less than 5% of the chosen sample actually return usable results.
::人们普遍认为,通过邮件、互联网或电话进行的任何调查的回复率都很低。 这种调查只有不到所选抽样的5%实际返回了可用结果,这是众所周知的。

播放段落

It certainly seems logical to attempt to get a response rate as high as possible, but does a low response mean that the experiment is invalid?
::试图获得尽可能高的回复率似乎是符合逻辑的,但低的回复率是否意味着实验无效?

播放段落

No, it doesn’t.  Certainly a very low response rate should be investigated, but the more important consideration is how well the results actually collected represent a random sampling of the population under study.
::当然,应该调查一个非常低的回复率,但更重要的考虑是,实际收集的结果在多大程度上代表了所研究人口的随机抽样。

播放段落

Examples
::实例

播放段落

Example 1
::例1

播放段落

What is the response rate of a survey of 368 persons, if 312 responses are collected?
::如果收集了312份答复,对368人的调查答复率是多少?

播放段落

$\begin{align*}\text{Response Rate} = \frac{(Number \ of \ responses \ collected)}{Number \ of \ responses \ attempted}=\frac{312}{368}=.847 \ or \ \approx 85\% \end{align*}$
::答复率=(所收到的答复数目)

播放段落

Example 2
::例2

播放段落

How many responses would you need to collect to achieve an 82% response rate from a survey of 1250 people?
::从对1 250人的调查中得出82%的答复率,你需要收集多少答复才能达到82%的答复率?

播放段落

$\begin{align*}0.82=\frac{Number \ of \ responses \ collected}{1250} \rightarrow 0.82 \times 1250 =Number \ of \ responses \ collected \rightarrow 1025\end{align*}$
::0.82=所收集答复数目1 2500.0.82x1250=所收集答复数目1025

播放段落

Example 3
::例3

播放段落

Is 428 responses sufficient to achieve an 80% response rate from a sample of 475 persons?
::475人的抽样答复率达到80%,428份答复是否足以达到80%?

播放段落

$\begin{align*}\frac{428}{475}=.90 \ or \ 90 \%\rightarrow Yes~ it~ is ~enough\end{align*}$
::428-4475=90或90________________________________________

播放段落

Example 4
::例4

播放段落

What range of responses would you need to achieve a response rate between 70% and 75% from a survey of 24,000?
::在对24 000人的调查中,要达到70%至75%的答复率,你需要什么答复范围?

播放段落

$\begin{align*}70 \times 24,000 =16,800 \end{align*}$ and   $\begin{align*}.75 \times 24000=18,000\end{align*}$ .  You would need between 16,800 and 18,000 responses.
::70×24,000=16,800 和 .75×24000=18,000 你需要16,800到18,000个答复。

播放段落

Review 
::回顾

播放段落

Look at the following table taken from a report located on EDU Cause's website.
::以下表格摘自EDU事业网站的报告。

播放段落
1. How many people total were polled?
   ::有多少人被民意调查?
播放段落
2. How many responded in the first go around?  What response rate does this represent?
   ::最初有多少答复?这代表什么答复率?
播放段落
3. How many responded in the second go around? What response rate does this represent?
   ::第二圈有多少答复?这代表什么答复率?
播放段落
4. What is the overall response rate?
   ::总的答复率是多少?
播放段落
5. Did the response rate appear to matter, if you compare column 1 and column 2 responses and percentages?
   ::如果比较第1栏和第2栏的答复和百分比,答复率似乎重要吗?
播放段落
6. What are the satisfaction ratings, should the data from the two columns be combined?
   播放段落
   1. Very Dissatisfied and dissatisfied:
      ::非常不满意和不满:
   播放段落
   2. Neutral:
      ::中立 :
   播放段落
   3. Satisfied and Very Satisfied:
      ::满意而且非常满意:
   
   ::满意度评分是多少? 这两栏的数据应该合并吗? 十分不满和不满:中立:满意和满意:满意和满意:
播放段落
7. Should there be an additional column representing the population that did not respond?  Create a chart showing percentages based on total population polled including non-responsive people polled.
   播放段落
   1. Very Dissatisfied and Dissatisfied:
      ::非常不满意和不满意:
   播放段落
   2. Neutral:
      ::中立 :
   播放段落
   3. Satisfied and Very Satisfied:
      ::满意而且非常满意:
   播放段落
   4. No Response:
      ::无答复:
   
   ::是否应该增加一列代表没有答复的人口的栏目? 创建一张图表,显示根据被调查的总人口的百分比,包括没有回应的被调查者。 非常不满和不满:中立:满意和非常满意:没有回应:
播放段落
8. “One type of survey error arises from what is known as nonresponse bias . This is the idea that what one learns from those who did respond may differ significantly from what might be learned from those who did not respond.” Based on this quote and the intention of the survey, what are some reasons that so many people may have been unresponsive?  What do you think the high percentage of “No Response” could tell us in this study?
   ::“一种调查错误产生于所谓的不答复偏差,这就是人们从答复者那里学到的教训可能与从没有答复者那里学到的教训大相径庭的想法。” 根据这一引文和调查的意图,为什么这么多人可能没有反应?你认为高比例的“无答复”在本研究报告中能告诉我们什么?
播放段落
9. Based on data gathered in question 6 and question 7, which information in this scenario do you think would be more important to the campus that conducted the research?
   ::根据在问题6和问题7中收集的数据,在这种假设中,哪些信息对进行研究的校园更为重要?

播放段落

Use the Graph below to answer the questions that follow.
::使用下图回答下列问题。

播放段落
10. What increases as population size increases regardless of the percent of responding population?
    ::不论答复人口的百分比如何,随着人口规模的增加,增加多少?
播放段落
11. If you expect to have a low response rate, what do you need to increase in order to improve the accuracy of your study?
    ::如果预计答复率较低,需要增加什么才能提高研究的准确性?

播放段落

An example of how to use the chart to determine accuracy:
::如何使用图表确定准确性的例子:

播放段落

You have a population of 1000, and you were able to contact via phone 500 of those people. You were able to get responses out of half of those you actually spoke to. So 25% of the total population participated in your study.  Using the chart above, find the intersection of 1000 on the horizontal axis and 25% on the vertical axis.  You would be approximately 95% certain of +/- 5% accuracy in your survey results.
::您有1000个人口,您能够通过电话与其中500个人口联系。您能够从您实际交谈的人中的一半得到回应。因此,25%的总人口参加了您的研究。使用上面的图表,在水平轴上找到1000个交叉点,在垂直轴上找到25%。您在调查结果中大约95%的精确度为+/-5%。

播放段落
12. You own a coffee shop.  You have about 750 regulars that come in and drink coffee at your coffee shop. You want to know if you should increase the speed of your wifi connection.  500 of your customers agree to take the survey home and complete it.  Only 300 of the customers returned their survey.
    播放段落
    1. What is your population size?
       ::你的人口规模是多少?
    播放段落
    2. What is your sample size?
       ::你的样本大小是多少?
    播放段落
    3. Based on participation, what can you expect the accuracy of your findings to be?
       ::根据参与情况,你对你的调查结论的准确性有什么期望?
    
    ::您拥有一家咖啡店。 您的咖啡店有大约750个常客, 来咖啡店喝咖啡。 您想知道您是否应该加快您的wifi连接速度。 500个客户同意将调查带回家完成。 只有300个客户回访了他们的调查。 您的人口规模是多少? 您的样本大小是多少? 根据参与情况, 您对调查结果的准确性有什么期望?
播放段落
13. As the manager of a ski resort, you are always looking for new ways to improve your customer’s experience.  You are thinking of adding an inner tubing/sledding slope right outside your hotel.  At peak season, you have 2000 people staying in the hotel.  You ask everyone who checks in to please complete the survey, but 356 people decline.  At the end of the survey period, you have received 693 surveys. 
    
    播放段落
    1. What is your population size?
       ::你的人口规模是多少?
    播放段落
    2. What is your sample size?
       ::你的样本大小是多少?
    播放段落
    3. Based on participation, what can you expect the accuracy of your findings to be?
       ::根据参与情况,你对你的调查结论的准确性有什么期望?
    
    ::作为滑雪度假胜地的经理,你总是在寻找改善客户经验的新途径。你正在考虑在旅馆外面添加一个内管/滑坡。在高峰季节,你有2000人住在旅馆里。你要求所有到访的人来完成调查,但有356人下降。在调查期结束时,你收到了693份调查。你的人口规模是多少?根据参与情况,你能期望你调查结果的准确性是什么?

播放段落

Standing outside a mall that reports a minimum of 500 people per day walking through the doors, you attempt to complete some research for school.  Your report will require an accuracy of +/- 10%.
::在商场外面站着,每天至少报告有500人走出门,你试图完成一些学校研究。你的报告需要10%的精确度。

播放段落
14. How many of the people walking through the door will you need to respond in order to achieve this accuracy?
    ::为了达到这一准确度,你需要多少人通过大门作出反应?
播放段落
15. If we also know that we need a 25% response rate, what would our sample size need to be?
    ::如果我们也知道我们需要25%的答复率,那么我们的抽样规模需要多少?

播放段落

Review (Answers)
::回顾(答复)

播放段落

Click to see the answer key or go to the Table of Contents and click on the Answer Key under the 'Other Versions' option.
::单击可查看答题键, 或转到目录中, 单击“ 其他版本” 选项下的答题键 。