How can we do a good job of questionnaire survey and explain all links?

First, the principles of questionnaire design

1. has a clear theme. According to the investigation theme, the topic is drawn up from reality, with clear purpose and prominent focus, and there are no redundant problems. 2. Reasonable structure and strong logic. The arrangement of questions should have a certain logical order, which conforms to the respondent's thinking program. Generally speaking, it is easy before difficult, simple after complex, concrete after abstract. 3. It's easy to understand. The questionnaire should make the respondents clear at a glance and be willing to answer truthfully. The tone of the questionnaire should be friendly, in line with the understanding and cognitive ability of the respondents, and avoid using technical terms. Take some skills to investigate sensitive questions, so that the questionnaire can be answered reasonably, avoid subjectivity and suggestion, and avoid the distortion of answers. 4. Control the length of the questionnaire. The time to answer the questionnaire is controlled at about 20 minutes, which neither wastes a question nor misses a question in the questionnaire. 5. It is convenient for data inspection, collation and statistics.

Edit paragraph 2. The procedure of questionnaire design.

The procedure of questionnaire design includes the following steps: 1: Grasping the purpose and content of the survey The first step of questionnaire design is to grasp the purpose and content of the survey, and the essence of this step is actually to specify the information needed to design the questionnaire. This is also the first step of scheme design. For researchers who are directly involved in the design of the research scheme, they can also skip this step and start from the second step of questionnaire design. However, for those researchers who have never participated in the scheme design, the first task when designing the questionnaire is to fully understand the purpose and content of this survey. Therefore, it is necessary to seriously discuss the purpose, theme and theoretical hypothesis of the research, read the research scheme carefully, consult and discuss with the scheme designer, and make the problem concrete, organized and operable, that is, turn it into a series of measurable variables or indicators. 2. Collect information about the research topic. Questionnaire design is not a simple imagination. In order to design the questionnaire perfectly, researchers need to know more. Questionnaire design is a technology that needs experience and wisdom. It lacks theory, because there is no scientific principle to ensure the best or ideal questionnaire. Questionnaire design is an art, not a science. Although there are some rules to follow to avoid mistakes, good questionnaire design mainly comes from the creativity of skilled researchers. There are three main purposes to collect relevant information: first, to help researchers deepen their understanding of the problems investigated; The second is to provide rich materials for problem design; The third is to form a clear concept of overall goal. Interviewing individual interviewees when collecting information is helpful to understand their experiences, habits, education level and understanding of questionnaire questions. We know very well that the problems that apply to college students may not necessarily apply to housewives. The greater the group difference of the respondents, the more difficult it is to design a questionnaire suitable for the whole group. 3. Determine the types of investigation methods. Different types of survey methods have influence on questionnaire design. In face-to-face investigation, respondents can see the problems and talk face to face with the investigators, so they can ask long and complicated questions of various types. In the telephone interview, the respondents can talk to the investigators, but they can't see the questionnaire, which determines that they can only ask short questions. The mailing questionnaire is filled out by oneself, and there is no direct communication between the respondents and the researchers, so the questions should be simpler and detailed. In computer-aided access (CAPI and CATI), we can realize more complex jumping and random arrangement problems to reduce the deviation caused by sequences. Questionnaires for personnel interviews and telephone interviews should be designed as dialogues. 4. Determine the content of each question Once the type of access is determined, the next step is to determine the content of each question: what should each question include, what should the questionnaire ask, whether it is comprehensive and relevant. Here, for each question, we have to ask: (1) Is this question necessary? (2) Do you need a few questions or just one? Our principle is that each question in the questionnaire should help to provide needed information or serve a specific purpose. If you can't get satisfactory usage data from a question, then the question should be cancelled. Of course, sometimes, you can "deliberately" ask questions and answers that are not directly related to the information you need. For example, asking some neutral questions at the beginning of the questionnaire can make the respondents willing to intervene and establish friendly relations, especially when the topic of the questionnaire is sensitive or controversial. Sometimes, some questions are "filled in" to cover up the purpose of the investigation or the funding (or entrusting) unit of the project. The first principle is to determine the necessity of a question, and then the second principle is to confirm the sufficiency of this question to the information obtained. Sometimes, in order to get the information you need clearly, you need to ask several questions at the same time. For example, the question of "why" in the choice of goods and schemes mostly involves two aspects: (1) "Not good-looking, but comfortable", (2) "Not comfortable, but good-looking" and (3) "Not good-looking and uncomfortable". Here, in order to get the required information, two different questions should be asked: (1) "Do you think XXX brand clothes look good?" (2) "Do you think XX brand clothes are comfortable to wear?" When determining the content of each question, researchers should not assume that the respondent can provide accurate or reasonable answers to all questions, nor should they assume that he will be willing to answer every known question. Investigators should try their best to avoid questions that respondents are "unable to answer" or "unwilling to answer". The appearance of "unable to answer" may be that the respondents "don't know", "don't remember" or "can't express". In the case of "not knowing", ask some "filtering questions" before asking, that is, measure past experience and familiarity, so as to filter out those respondents who don't know the situation. Secondly, respondents may not remember some of the survey contents. The results show that the ability to recall an event is influenced by three factors: (1) the event itself; (2) the time span of the event; (3) Whether there are other events that may be helpful to memory. The questions reviewed in the questionnaire may be useless or beneficial. Useless memories generally lead to underestimation of the actual situation. For example, respondents are asked to answer a question and answer the question without any hint: "What products did you watch on TV last week?" This is an example of useless memory. And if you list the names of a series of products or enterprises and ask, "Which of the following enterprises or products did you see advertisements last week?" This is a helpful memory, which can stimulate the memory of the respondent by giving some hints. Respondents are unable to express their answers to some types of questions. For example, it is often difficult to express accurately when respondents ask what atmosphere they like to eat in restaurants. However, if some alternative answers are given to describe the hotel atmosphere, respondents can point out which one they like best. Otherwise, if they can't express themselves, they may ignore this question and refuse to answer the rest of the questionnaire. So we should provide some help, such as pictures, maps and descriptive words, to help them answer. There are several situations in which respondents are "unwilling to answer". One is that respondents must try to provide information; Secondly, some questions in the survey are inconsistent with the background of the survey (for example, the consumption of ordinary goods and personal privacy are not consistent in the same questionnaire); Third, the reasonable purpose of the investigation, without which the respondents are unwilling to provide the investigation; The fourth is a sensitive issue. There are several ways to encourage respondents to provide unwanted information: (1) Put sensitive questions at the end of the questionnaire. At this point, the respondents' alertness has been greatly weakened and they are willing to provide information. (2) Add a "preface" to the question and answer, explaining the background and * * * situation of relevant questions (especially sensitive questions)-to overcome the respondent's fear that his behavior does not conform to social norms. (3) Using "third party" technology to ask and answer questions, that is, to intervene in the question from the perspective of others. 5. Deciding the structure of questions and answering questions Generally speaking, there are two types of questions in the questionnaire: closed questions and open questions. Open-ended questions, also known as unstructured questions and answers, are freely answered by the respondents in their own language, without providing specific questions for choosing answers. For example, "Why do you like Nike TV commercials?" "What do you think of the current state-owned enterprise system reform in China?" Open-ended questions allow respondents to fully express their views and reasons, and they go deeper and sometimes get unexpected answers from researchers. Its disadvantages are: there is a lot of useless information in the collected materials, which is difficult to make statistical analysis, and the records of face-to-face interviews of investigators directly affect the survey results, and may be rejected because of troublesome answers. Therefore, open questions are very helpful in exploratory investigation, but they do more harm than good in large-scale sampling investigation. Closed question and answer, also known as structured question and answer, stipulates a set of alternative answers and a fixed answer format. For example, what are the main factors you consider when choosing to buy a house? (a) Price (b) Area (c) Traffic conditions (d) Surrounding environment (e) Design (f) Construction quality (g) Other _ _ _ _ _ _ _ _ _ _ _ _ (Please specify) Advantages of closed-ended questions include the following aspects: (1) The answer is. (2) Respondents are easy to answer, which is conducive to improving the questionnaire recovery rate; (3) The meaning of the question is clear. Because the answers provided help to understand the meaning of the question, it can prevent the respondents from refusing to answer because they don't understand the meaning of the question. Closed-ended questions also have some shortcomings: (1) it is difficult to detect what the respondents did not understand correctly; (2) There may be "order deviation" or "position deviation", that is, the answers selected by the respondents may be related to the arrangement position of the answers. Research shows that respondents tend to choose the first or last answer, especially the first answer, for declarative answers. For a set of numbers (quantity or price), tend to take the middle position. In order to reduce the order deviation, several forms of questionnaires can be prepared, and the answers of each form are arranged in different order. 6. Decide on the wording of the question. Arrange the order of questions. Determine the format and layout. Draft questionnaire, pre-survey 10. Make a formal questionnaire.

Edit paragraph 3. The form of answering questions

1. Open-ended questions are also called unstructured questions and answers. When using open-ended questions, respondents can express their opinions freely in their own language, and there are no prepared answers in the questionnaire. How long have you been smoking? What kind of TV programs do you like to watch? What impact do you think China's entry into WTO will have on the government management system? Obviously, respondents can freely answer the above questions, and do not need to choose according to the answers already drawn up in the questionnaire, so respondents can fully express their views and reasons, and sometimes they can get unexpected answers if they go deeper. Generally speaking, the first question in the questionnaire is a free-form question, so as to give the respondents the opportunity to express their opinions as much as possible, thus creating a favorable investigation atmosphere and shortening the distance between the investigators and the respondents. However, the open question also has its disadvantages. For example, the investigator's bias, because the respondent's answer was written by the investigator, is likely to be distorted or not the respondent's intention. If the investigator records according to his own understanding, there is the possibility of deviation. But these shortcomings can be made up by using a tape recorder. The second main disadvantage of open-ended problems is the difficulty of data sorting and analysis. Because each respondent's answers may be different and the words used are different, it is inevitable that there will be difficulties in the classification of answers, and the whole process is quite time-consuming, and it is inevitable that the organizer's personal prejudice will be mixed. Therefore, open questions are very helpful in exploratory investigation, but they do more harm than good in large-scale sampling investigation. 2. Closed questions are also called structured questions and answers. Closed questions, contrary to open questions, stipulate a set of alternative answers and a fixed answer format. For example, the main reason why you buy Di 'ao brand washing powder is (choose the two most important ones): (1). Clothes are whiter (2). The price is cheaper (3). They are sold in any store. Do not hurt hands (5). The price is the same as the existing brand, but the weight is more (6). The advantages of closed-ended questions introduced by friends include the following aspects: (. (2) Respondents are easy to answer, which is conducive to improving the questionnaire recovery rate; (3) The meaning of the question is clear. Because the answers provided help to understand the meaning of the question, it can prevent the respondents from refusing to answer because they don't understand the meaning of the question. Closed questions also have some disadvantages: (1). It is difficult to detect what the respondent did not understand correctly; (2) There may be "order deviation" or "position deviation", that is, the answers selected by the respondents may be related to the arrangement position of the answers. Research shows that respondents tend to choose the first or last answer, especially the first answer, for declarative answers. For a set of numbers (quantity or price), tend to take the middle position. In order to reduce the order deviation, several forms of questionnaires can be prepared, and the answers of each form are arranged in different order. 3. The scale response problem is set in the form of scale.

Edit the fourth paragraph. Questionnaire type

The types of questionnaires can be divided from different angles. According to the answer to the question, it can be divided into three types: structured, semi-structured and open; According to the investigation methods, it can be divided into interview questionnaire and self-filled questionnaire; According to the purpose of the questionnaire, it is divided into screening questionnaire, questionnaire and return visit questionnaire. First of all, according to the answer to the question. Questionnaire can be divided into three basic types: structured, open and semi-structured. 1. structural formula. Also known as off or off. The answer of this questionnaire has been determined by the researcher on the questionnaire, and it is enough for the respondent to carefully choose an answer and circle or tick it. 2. open it. Also known as open. This kind of sentence paper has no fixed answer, so that returnees can play freely. 3. Semi-structural structure. This kind of questionnaire is between structured and open. The answers to the questions are both fixed and standard, and some of them are freely played by the respondents, drawing on the advantages of both. This kind of questionnaire is widely used in actual investigation. Second, according to the survey method. According to the survey methods, questionnaires can be divided into: self-filled questionnaires and interview questionnaires. Self-administered questionnaire is a questionnaire filled and answered by the respondents themselves. The interview questionnaire is a questionnaire that the interviewer answers by interviewing. Due to different sending methods, self-filled questionnaires can be divided into two categories: sending intermediate papers and mailing questionnaires. Sending a questionnaire is a survey form in which the investigator directly sends the questionnaire to the interviewed voice and the investigator directly recovers it. Mailing questionnaire is a form of investigation that is directly mailed by the investigation unit to the respondents, who fill in the answers themselves and then mail them back to the investigation unit. The characteristics of these forms of investigation are as follows: the interview questionnaire has the highest recovery rate and the most reliable results, but it is costly and takes a long time, and the recovery rate of this kind of indirect paper is generally required to be above 90%; The outgoing text and incoming string are low, the investigation process is uncontrollable, and the credibility and effectiveness are low. Moreover, due to the low recovery rate, the sample will be biased, which will affect the report of the sample to the whole. Generally speaking, the recovery rate of mailing questionnaires is about 50%; The advantages and disadvantages of sending a self-filled questionnaire are between the above two, and the recovery rate is required to be above 67%. Third, according to the purpose of the questionnaire. Generally speaking, questionnaire survey, especially market survey, includes three types of questionnaires, namely screening questionnaire, questionnaire and return visit questionnaire (review questionnaire). 1. Screening questionnaire. The screening questionnaire is a set of questions to ensure that the respondents in the survey are indeed the target consumer voices of the survey products. Generally, it includes five aspects: the exclusion of individual natural state variables, the exclusion of product applicability, the exclusion of product use frequency, the exclusion of States that have special influence on product evaluation and the exclusion of investigation refusal.