The percentage of questionnaire shortages cannot be used for analysis

Is the rate important? I send out 100,000 questionnaires, **10,000, **rate 10% You send out 50 questionnaires, **50 questionnaires, **rate 100%** questionnaire is enough to sample the maternal needs.

The main thing is to look at the effectiveness of the questionnaire. If the effective rate of the questionnaire is less than 90%, it means that our research method is problematic and the survey data has lost its value.

Questionnaires can be divided into: self-administered questionnaires and interview questionnaires. A self-administered questionnaire is a questionnaire that the respondent fills out by himself/herself.

The interview questionnaire is a questionnaire filled out by the interviewer through the interviewee's voice. Self-administered questionnaires are divided into two categories: sending questionnaires and mailing questionnaires due to different delivery methods. The intermittent questionnaire is a form of investigation in which the questionnaire is directly sent to the interviewee by the investigator and directly sent by the investigator.

The mailing questionnaire is a survey form in which the survey unit directly mails the respondents, and the respondents fill in the answers by themselves, and then mail them to the investigation unit.

In general, the ** rate of the interview questionnaire is the highest, and the ** rate is generally required to be above 90%;

The **rate of mailing questionnaires is low, and the **rate is about 50%;

The ** rate of sending self-administered questionnaires is average, and the ** rate is required to be more than 67%.

There is no fixed quantity, it depends on the **rate, and the **rate of 90% is enough. In the process of investigation and implementation, it is necessary to check the data in a timely manner, check whether the collected information is complete and accurate, pay attention to whether the figures in the data are correct, and make timely remedies and corrections, so as to avoid human errors as much as possible.

The main content is to tell people what work to do during and after the implementation of the questionnaire to ensure that the information filled in the survey is complete, correct and effective. After the questionnaire, after the elimination of the waste papers, the rate of effective questionnaires should be counted.

There is no specific requirement for the number of questionnaires. For particularly large projects, a larger number may be required; And for some small projects, there is naturally less need. This range is as many as hundreds of thousands and as few as a few dozens.

As for the effectiveness, he looks at it according to the **rate, and the ** is naturally higher; But that's not exactly the case, because in the end your data is based on the questionnaire that came back.

If you want more accurate data, you need to pay more for a survey, and if the data you need is only a rough calculation, you don't have to make such a hard copy. I think it's more modest than 700 copies (only in Beijing, including the suburbs).

There is no clear definition of this range!

In principle.

The more the more accurate (in the case of guaranteed ** rate).

A recent CCTV questionnaire survey had a score of 100,000.

**81,200 servings.

The highest rate!

This problem can be solved by stratified sampling in mathematical sampling surveys. This approach mainly addresses surveys with a large overall population. Let's say you want to investigate a situation in a secondary school.

The first thing to do is to determine what the number of people you want to actually investigate. For example, n1. Find out the total number of students in the target school n2.

Then the target school is stratified according to the law, such as according to the grade (can also be divided according to other methods), divided into 6 levels, the number of people is a, b, c, d, e, then the number of people who should be investigated in the first year of junior high school is a*n1 n2, and so on, to determine the number of people in each layer can be randomly selected separately, and then recorded.

Generally, the larger the n1 take, the higher the accuracy, and of course, the more work.

The non-response rate should be looked at from many aspects.

1. The non-reply rate caused by the survey method and process has no impact on the accuracy 2. The non-reply rate of the survey items, that is, the non-reply rate caused by the questionnaire questions, reflects that there is a problem in the preparation of the survey planning and the setting is unreasonable.

Generally speaking, the non-response rate reflects the status of the question setting, which affects the accuracy of the results, but if the non-response rate is high, it indicates that there is a problem with the survey setting, and it also reflects the one-sided tendency of the survey results.

Finally, if it is 1 above, the non-response rate caused by it is low, and it does not affect the accuracy of the survey, if it is 2, the non-response rate caused by it cannot be higher than 40%, otherwise it is not representative.

Looking at your total sample size, and your statistical methods, some methods don't require large sample sizes.

SPSS method of analyzing questionnaire data.

It is necessary to take into account the convenience of statistics in the design in order to facilitate the summary. Excel will do.

I have a good grasp of the number of SPSS that I can analyze on behalf of and have years of experience analyzing data for graduate students.

The questionnaire should have more than 20 questions, and no more than 30 is the most appropriate.

In the design of these 20 or so questions, we want to ensure that they are not repetitive, and that each question should be independent. It is also necessary to include comprehensive survey information in the design of the questionnaire questions, and a good questionnaire is not only concise and clear, but also has a strong general content.

The questions of the questionnaire must be determined according to the content of the survey and the purpose of the survey, and in general, the detailed questions are determined according to the specific content and purpose of the survey, as well as the comprehensive factors such as the respondents.

The content and number of questions in the questionnaire are affected by the above comprehensive factors, but in general, the questionnaire questions should not be too much, and it is good to be simple and complex, for example, the questionnaire has ten questions, so it is best to have single-choice, multiple-choice and free-play questions, so that the respondent will have patience and interest to fill in.

The general questionnaire should not have more than 20 questions, if it is for professionals and people in specialized industries, the questionnaire can have more questions, if it is an ordinary questionnaire, the questions are about fifteen, for example, the questionnaire of the street survey is suitable for fewer questions.

The number of questions in the questionnaire should be controlled, and too long the questionnaire will bore the participants and reduce the reliability and validity. In addition, when doing the questionnaire, give a small gift to the participant, and the appropriate dress, verbal expression, attitude and similar age of the respondent can get the favor of the participant and get better cooperation.

Judging from the several questionnaires we have done, if it is an internal survey conducted by the company, the number of questions can be a little more, but it cannot be unlimited, and at the same time, according to the content of the questionnaire, it is generally set within 50 questions.

If you are conducting a questionnaire survey on society, it is even more important to reduce the number of questions, and experience tells us that 30 questions will bore the respondents.

If you do a ** questionnaire, it is better to have less than 10 questions, and I feel that 3 or 5 questions are the best.

There is no fixed number of questions to be set in the questionnaire, but it is determined according to the purpose of the survey.

If the purpose is relatively simple, you may set up a question, such as often receiving ** reviews from mobile customer service, "You rate our service". This is a survey of questions in order to understand the satisfaction of an overall service;

If the purpose is more responsible, such as to understand the consumer's consumption behavior and habits, which is used for market competition analysis, and involves all aspects such as the consumer's use of this product and the use of competing products, there will be many such questionnaire questions.

Some market research companies will also conduct ride-hailing surveys, which are several different purposes, and multiple questionnaires are combined to conduct surveys, up to dozens of pages of questionnaires.

However, considering the receptivity of the respondents, if there are many questions that are not suitable for interception and interviews on the street, it is necessary to enter the house or invite to fill in the answers. If there are few questions, you can ** or block access.

She generally controls the questionnaire to about 20, and no more than 30 is the most appropriate.

Difference studies usually include the following types of analysis methods, which are:ANOVA, t-test, and chi-square test。The similarities and differences between these three analysis methods are as follows:

In fact, the core difference is that the data types are different. In the case of categorical and categorical classification, chi-square analysis should be used at this time; If it is categorical and quantitative, variance or t-test should be used.

The difference between the variance and the t-test is that for the x of the t-test, it can only be 2 categories, such as male and female. If x is 3 categories, such as undergraduate, undergraduate, undergraduate; Only ANOVA can be used in this case.

t-test, ANOVA, chi-square test, etc.

This is to do a questionnaire to find out why the answers to the questionnaire are different.