Questionnaire survey is a survey method widely used in the survey industry at present - that is, the survey agency designs various questionnaires according to the purpose of the survey, and then adopts a sampling method (random sampling or cluster sampling) to determine Survey samples are a way for investigators to interview the samples and complete pre-designed survey items. Finally, statistical analysis is used to obtain survey results. It strictly follows the principles of probability and statistics. Therefore, the investigation method is highly scientific and easy to operate. The impact of this method on survey results depends on factors such as sample selection, investigator quality, and statistical methods. The level of questionnaire design is one of the prerequisites. The quality of questionnaire design is largely related to the design system (principle)!
1. Rationality. Reasonability means that the questionnaire must be closely related to the topic of the survey. If it violates this point, no matter how beautiful or exquisite the questionnaire is, it will be useless. The essence of the questionnaire reflecting the survey theme is to find out the "elements related to the survey theme" at the beginning of the questionnaire design!
For example: "Investigating users' consumption experience of a certain cosmetic product" - there is no ready-made rule for selecting factors here. But starting from the problem, especially when combined with certain industry experience and business knowledge, the elements can be found: First, the user (can be identified as the buyer). Including her (his) basic situation (natural conditions: such as gender, age, skin properties, etc.); the use of cosmetics (whether the cosmetics have been used, cycle, daily habits of using cosmetic crystals, etc.); second, purchasing power and desire to buy . Including her (his) social status, income level, education level, occupation, etc.); cosmetics consumption characteristics (brand, packaging, price, product appearance, etc.); the effect of using the cosmetic crystals (evaluation. Questions should have a certain diversity , but limited to a certain range, such as 1. Price; 2. Use effect; 3. Psychological satisfaction, etc.); The third is the product itself. Including evaluation of packaging and trademarks, the influence of promotional methods such as advertising, horizontal comparison with similar products on the market, etc... It should be said that having these elements will directly help the results of the investigation topic. It is also relatively easy for the interviewee to understand the investigator's intentions and thus cooperate.
2. Generality. That is, whether the problem setting has general significance.
It should be said that this is a basic requirement for questionnaire design, but we can still find such common-sense errors in questionnaires. This error is not only detrimental to the organization and analysis of investigation results, but also causes the investigation client to underestimate the level of the investigator. For example, conduct a survey on "residents' acceptance of advertising":
Question: Which advertising media do you usually choose:
Answer: a. Newspaper; b. TV; c. Magazine ;d. Broadcasting; e. Others
And if the answer is another form:
a. Newspaper; b. Ticket; c. TV; d. Wall advertising; e , balloon; f. big bus; 8. advertising shirt; h,...
If our statistical indicators are not so detailed (or not necessary at all), then we have committed a "speciality" Errors, resulting in answers to certain questions that are actually unhelpful to the investigation!
In general questionnaire techniques, what needs to be noted is: you cannot make mistakes in the content of the questions. For example:
Question: Which kind of credit card do you have?
Answer: a. Great Wall Card; b. Peony Card; c. Dragon Card; d. Visa Card; e. Gold Card Suica,
----The setting of "d" is wrong and should be avoided.
3. Logic. The design of the questionnaire must have a sense of integrity. This sense of integrity means that the questions must be logical, and independent questions themselves must not have logical fallacies. This makes the questionnaire a relatively complete small system.
For example:
Questions:
1. How many newspapers do you usually read every day?
a. No newspapers; b. 1 copy; c. 2 d. more than 3 copies;
2. How long do you usually spend reading newspapers?
a. Within 10 minutes; b. About half an hour; c. 1 hour; d. More than 1 hour;
3. Which of the following types (or categories) of newspapers do you often read?
a. × City Evening News; b. × Provincial Daily; c. , People's Daily; d. Reference news; e, China Central Radio and Television; f, Football...
Among the above questions, because the question settings are closely related, relatively complete information can be obtained. Respondents will also feel that the questions are focused and the questions are organized. On the contrary, if the questions are divergent and have traces of stream of consciousness, the questionnaire will give people a sense of randomness rather than rigor. Then, companies that regard market research as a scientific process for business decision-making will lose confidence in the survey!
Therefore, logical requirements are inseparable from the orderliness and procedural nature of the questionnaire. It has been seen that in a comprehensive questionnaire, the investigator divides the questionnaires with large differences into blocks, thereby ensuring that the questions in each "block" are closely related.
4. Clarity. The so-called clarity is actually the standardization of problem setting. This principle specifically refers to: Is the proposition accurate? Is the question clear and easy to answer? Is the interviewee able to give a clear answer to the question, etc.
For example, the design of "10 minutes", "half hour", "1 hour" in the above question is very clear. Statistics will tell us: What is the probability of a very short time (survey); what is the probability of an average time (rough reading); what is the probability of a long time (detailed reading). On the contrary, if the answer is set to "10 minutes to 60 minutes", or "within one hour", etc., it will not only be unclear and difficult to explain the problem, but also make it difficult for the interviewee to answer.
Furthermore, there are often yes-or-no propositions such as "yes" or "no" in questionnaires. For example:
Question: Your marital status:
Answer: I. Married; II. Unmarried;
Obviously, there is a third way to this question Answer (divorced/widowed/separated). If set up in the above way, difficulties in selection and loss of effective information will inevitably occur! The crux of the problem is that the questionnaire violates the principle of "clarity".
5. Non-inducing. Unsuccessful reporters often use leading questions in interviews. ----If this kind of questioning method is not deliberately trying to reach a certain conclusion but is willing to give up the principle of objectivity, it is a complete lack of professional quality. In questionnaire surveys, this kind of error is greatly reduced because there is sufficient time to prepare in advance. But the reason why this principle becomes necessary is that the highly competitive market has put forward higher requirements for the development of the investigation industry.
Non-inducing means that the questions should be set in a neutral position, without involving prompts or subjective assumptions, and the independence and objectivity of the interviewee should be completely placed in the position of limiting conditions for the operation of the questionnaire. For example:
Question: What do you think is the attraction of this cosmetic to you?
Answer: a. Color; b. Smell; c. Effect of use; d. Packaging; e. Price; f....
This setting is objective. If you change the answer setting:
a. Charming color; b. Fragrant smell; c. Satisfactory effect; d. Exquisite packaging...
Such a setting It is inducing and suggestive, thereby unconsciously concealing the authenticity of things.
6. Easy to organize and analyze. In addition to taking into account the close integration of survey topics and convenient information collection, successful questionnaire design also takes into account the ease of obtaining survey results and the persuasiveness of the survey results. This requires consideration of the sorting and analysis of the questionnaire after the survey.
First of all, this requires that the survey indicators can be accumulated and be easy to accumulate; secondly, the accumulation of indicators and the calculation of relative numbers are meaningful; thirdly, the issues to be investigated can be clearly explained through the data.
Only in this way can the investigation work achieve the desired results.
For more information, please go to: I Love Investigation Network