Statistics 8th Edition Jia Junping Chapter 2 after class exercises and answer summary

1 What is second-hand information ？ What should we pay attention to when using second-hand materials ？

answer ：（1） If the original information related to the research content already exists , Just reprocess these original information 、 Arrangement , Make it the data that can be used for statistical analysis , Data from indirect sources , That is, second-hand information .
  （2） Before using second-hand materials , It is necessary to evaluate second-hand data . The evaluation of second-hand data should consider the following ：

① Who collected the data ？ This is mainly to examine the strength and social credibility of data collectors .

② What is the purpose of the collection ？ The data collected for the benefit of a group is questionable .

③ How data is collected ？ There are many ways to collect data , Data collected by different methods , Its explanatory power and persuasive power are different . If you don't know the method used to collect data , It is difficult to make an objective evaluation of the quality of data . The quality of data depends on the production process of data .

④ When did you collect it ？ For outdated data , Its persuasiveness is naturally questioned .
in addition , Use second-hand data , Pay attention to the definition of data 、 meaning 、 Calculation caliber and calculation method , Avoid misuse 、 misuse 、 The abuse of . When quoting second-hand data , The source of the data should be indicated , In order to respect the achievements of others .

2. Compare the characteristics of probability sampling and non probability sampling . Give examples to illustrate when probability sampling is suitable , Under what circumstances is non probabilistic sampling suitable .

answer ：（1） Characteristics of probability sampling and non probability sampling

① Probability sampling is to select samples according to the random principle , At this time, the theoretical distribution of sample statistics exists , Therefore, the relevant parameters of the population can be estimated according to the results of the sample , Calculate the estimation error , The confidence interval of population parameters is obtained ; During sampling design , Put forward requirements for the accuracy of estimation , The required sample size can be calculated ; Probability sampling is more technical , Whether sampling or analyzing the survey data , All require more professional knowledge of Statistics , The cost of the survey is also higher than that of non probabilistic sampling .

② Non probabilistic sampling is not based on random sampling , The distribution of sample statistics is uncertain , Therefore, the results of samples cannot be used to infer the corresponding parameters of the population . Non probabilistic samples are characterized by simple operation 、 Fast time 、 The cost is low , Moreover, the professional and technical requirements for statistics in sampling are not very high .
（2） Where probability sampling and non probability sampling are applicable

① If the results of the survey require an estimate of the overall relevant parameters , Calculate the estimated error , The confidence interval of population parameters is obtained , And put forward requirements for the accuracy of estimation , Then probability sampling should be selected , Such as investigating the consumption level of different age groups .② Non probabilistic sampling is suitable for exploratory research , The results of the survey are used to identify problems , Prepare for more in-depth quantitative analysis . Non probabilistic sampling is also suitable for concept testing in Market Research , Such as product packaging test 、 Advertising test, etc , At this time, it is not necessary to project the survey results to the overall situation .

3 The main methods of collecting data in the survey are self filling 、 Face to face interview 、 Telephone . besides , What other ways to collect data ？ answer ： The method of collecting data except self filling 、 Face to face interview 、 Telephone , And observation , That is, investigators obtain information through direct observation , For example, use the video equipment installed in the supermarket to observe the expression of customers when choosing goods ; Estimate the traffic flow by counting at the intersection .4 Self filling 、 Face to face interview 、 What are the advantages and disadvantages of telephone surveys ？

answer ：（1） Self filling means that the respondent fills in the form without the assistance of the investigator , Complete the questionnaire

① Advantages of self filling ：
a． It is relatively easy for survey organizers to manage the self filling method , As long as the questionnaire is correctly delivered to the respondents ;b． The lowest cost of investigation , Increasing the sample size has little impact on the investigation cost , Therefore, a large-scale investigation can be carried out ;
c． Respondents can choose a convenient time to fill out the questionnaire , You can refer to relevant records instead of relying on memory to answer . This method can reduce the pressure of respondents to answer sensitive questions to a certain extent .

② Disadvantages of self filling ：
a． The recovery rate of the questionnaire is relatively low ;
b． The self-administered method is not suitable for questionnaires with complex structures , The content of the survey will be limited ;
c． The investigation cycle is usually long ;
d． For the problems in the process of data collection , It is generally difficult to take adjustment measures in time .
（2） Face to face interview means that the investigator and the respondent face to face in the on-site investigation , The investigator asked 、 The respondent answered

① Advantages of face-to-face interview ：
a． It can stimulate the respondents' awareness of participation , Improve the response rate of the survey ;
b． The investigator can explain the questionnaire on site , Answer the questions of the respondents , Identify and clarify the respondents' answers , Improve the quality of survey data , And we can investigate groups with low literacy rate ;
c． More technical means can be used in the questionnaire design , Make the combination of investigation questions more scientific 、 reasonable . for example , You can use other survey tools （ picture 、 Photo 、 card 、 Physical objects, etc ） Enrich the content of the survey ;
d． Be able to adjust the time spent in data collection .

② Disadvantages of face-to-face interview ：
a． The cost of investigation is relatively high ;
b． There are certain difficulties in the quality control of the investigation process ;
c． For sensitive issues , In face-to-face conditions , Respondents usually don't relax as they would under the self-administered method , Also unwilling to answer , There may also be counterfeiting .
（3） Telephone type refers to the way that investigators make a phone call to investigate the respondent

① The advantages of telephone ：
a． Fast , Be able to complete the survey in a very short time ;
b． It is especially suitable for the situation that the sample units are very scattered , Because there is no need to pay the investigator's transportation fee , The cost of data collection is relatively low ;c． It's safe for investigators ;
d． It is easy to control the access process .

② The disadvantage of telephone ：
a． Easily limited by telephone tools ;
b． The time for access cannot be too long ;
c． The questionnaire used in the survey should be concise , There can't be too many problems ;
d． Because it's not face-to-face communication , When the respondent is unwilling to be investigated , It will be more difficult to persuade them .5 Please cite （ Or design ） Several examples of experimental data . answer ： Experimental data refers to the data of variables collected by controlling the experimental object in the experiment . for example ： For a group of livestock raised together , Feed different feeds respectively , To test the effect of different feeds on livestock weight gain ; For several farmland with exactly the same other conditions , Apply different fertilizers respectively , To test the effect of different fertilizers on farmland harvest .

6 How do you think we should control the response error in the survey ？

answer ： Answer error refers to the fact that the answer given by the respondent during the survey is inconsistent with the real situation . To control the response error in the survey, we need to analyze the causes of the response error , There are many reasons for response errors , There are mainly understanding errors 、 Memory error and conscious error .
（1） For understanding errors , Pay attention to the wording in the questionnaire , Design a good questionnaire ;
（2） For memory error , The time interval involved in the investigation should be shortened ;（3） For conscious errors , Investigators should do a good job in the ideological work of the respondents , Let them dispel their concerns ; Investigators should abide by professional ethics , Keep secret for the investigated ; Try to avoid sensitive issues in the investigation .

7 How to reduce unanswered ？ Please pass an example , Explain the specific measures you are considering to reduce unanswered .

answer ： We should do a good job to prevent the error of no answer , There is no answer error , We should analyze the causes of errors in detail , To remedy .
（1） If no answer error is random , It can be solved by increasing the sample size .
for example ： The survey design requirements are completed 1000 Sample units , As a result, I answered 800 individual , The no answer rate is 20%, At this time, you can draw randomly 250 A unit of , And investigate it , If the no answer rate is still 20%, You can get 200 Answer of units . At the same time, investigators should also increase their sense of responsibility and patience .
（2） If there is no response error, it is a systematic error , The solution is mainly in two aspects ：① The prevention of , That is, make preparations in all aspects before the investigation , Try to minimize no answer . for example , To investigate the monthly living expenses of residents in a certain area , Where possible , We can make full use of the authority and social influence of the survey organizers , Stimulate the respondents' awareness of participation ; in addition , Notice in advance before investigation 、 Give a detailed explanation of the investigation ; Hire someone familiar with the respondent to be the investigator , Such as members of the residents' committee 、 Laid off workers also help dispel the concerns of the respondents .② When no answer appears , Analyze the cause of no answer , Take some remedial measures . for example ： Take another sample from the unanswered unit , Conduct stronger investigations , And take this as the representative of the unanswered layer , Combined with the data of the response layer to estimate the overall .