The 97-2003 edition seems to have been forgotten in the tool.
You use help to search,
Question 2: The statistical method used in the questionnaire survey is 50 points 1. The sample size of the investigation is too small, and the calculated conclusion is unreliable.
For example, when I saw some graduate papers, I only sent dozens of questionnaires and wrote a very positive conclusion according to the statistical percentage. Actually, there is a problem.
For example, 45 people participated in the survey "How much do you like XXX activities". Survey results: I like 2 people very much, like 5 people, generally 10 people, dislike 13 people, and dislike 15 people. According to the author's statistics, 7 people like * * * very much, accounting for 15.5% of the 45 people surveyed, and 28 people don't like * * *, accounting for 62.2%. And further according to 15.5% and 62.2% to write the conclusion.
However, after calculating the ratio from the survey sample, he ignored the standard error and confidence interval of calculating the ratio. For example, the like rate of this example is 15.5%. You should also calculate the standard error Sp of the ratio.
_________ _________________
In this example, the standard error of liking rate is sp = √ p (1-p)/n = √15.5 (100-15.5)/45 = 5.39%.
According to the sample size n, check the values of t0.0 1 and t0.05 of n- 1, and get t0.05 = 2.02, T0.0 1 = 2.69. According to 15.5% similarity, 5.39% standard error and T0.09,
95% confidence interval:15.52.02× 5.39 = 4.6% ~ 26.4%. (The difference between the upper and lower limits of the confidence interval is as high as 2 1.8%).
95% confidence interval means that if the sample's like rate is 15.5% to estimate the overall like rate, 95% may be between 4.6% and 26.4%. Such an interval as high as 2 1.8% means that 15.5% is not credible.
However, if the sample size is expanded to 450 and 4500 people, the statistical liking rate is also 15.5%. With the expansion of the sample size, the standard deviation Sp will decrease, and the calculated 95% confidence interval will be reduced to 12.2% ~ 18.8% and 14.4% ~ 16.6%. At this time, when the sampling rate is used to estimate the population rate, the difference between the upper and lower limits is very close to 15.5%, which is credible.
2. The statistical analysis of survey data is too simple.
At present, the statistical analysis of survey data is mostly simple. Just calculate the percentage of each questionnaire indicator, such as the above-mentioned 15.5% like rate.
To avoid too simple statistical analysis, we should first consider the statistical analysis method of survey data in advance when designing the questionnaire. For example, it is also a survey of "How much do you like XX activities". In addition to expanding the sample size of the survey, gender and age were added to the questionnaire. In this way, a more complicated method-cross analysis method can be adopted. Cross-analysis is to analyze the relationship between three variables: age, gender and XXX activities. Assuming that statistics are not classified, the like rate is 15.5%. After cross-analysis, it will be found that due to gender differences, different age groups have different liking rates.
Take the questionnaire survey of national physique monitoring in 2005 as an example. If we simply calculate the questionnaire of 2473 adult men in a city, we can only get statistical data: those who sleep less than 6 hours 13.4%, those who sleep for 6-9 hours/73.6% and those who sleep for more than 9 hours 13%. But if we add the age factor, we can see that the percentage of each age group is different (the statistical table is omitted). You can also draw a line chart (abbreviated) with the percentage of age group. It is obvious from the picture that with the increase of age, the sleep time decreases gradually.
The above statistical analysis method is relatively simple. However, it is a pity to only use simple statistical methods to process the questionnaire data, because a large amount of data information has not been fully utilized. Therefore, when designing the questionnaire, it should be noted that the collected survey data can be used for multivariate statistical analysis (such as regression analysis and factor analysis). The following are examples of statistical analysis that I have done to help or guide relevant units:
Example1:The contents of the national physique monitoring questionnaire in 2005 included many questions about everyone's education level, occupation, work, life, physical exercise and so on. In order to analyze the relationship between these surveys and everyone's physique and find out which factors are beneficial to physique ...
Question 3: How to make a statistical analysis of the questionnaire Eviews software is easy to use. Generally, statistical analysis with excel is enough. In fact, it doesn't matter if the method is simple and complicated, as long as it is suitable.
Question 4: How to make a statistical analysis of the questionnaire Eviews software is easy to use. Generally, statistical analysis with excel is enough. In fact, it doesn't matter if the method is simple and complicated, as long as it is suitable.
Question 5: How to analyze questionnaire data with Excel?
Among them, a new table named "Summary" is created as a template, then this table is copied and renamed as 1, the first questionnaire result is entered, and then another table is copied and the second questionnaire result is entered. . . Until the input is complete.
Then enter the summation formula in the summary table.
B2 formula is as follows:
= Su Wa ('Summary (2): Summary (4)'! B3)
Abstract (2) is the name of the first questionnaire result table, and abstract (4) is the name of the last questionnaire table. For simplicity, I made three result tables, and then copied the formula into all the cells.
Question 6: How to analyze the data after entering a good questionnaire? When designing, we should consider the convenience of statistics to facilitate summary. Just use excel.
Question 7: Urgent ~ How to make a statistical analysis of the questionnaire survey? We can compare respondents with different characteristics.
I do a lot of data analysis for others.
There are countless kinds of tools, and spss is the most commonly used.
Question 8: How to analyze the questionnaire 5 points If you analyze a specific chicken questionnaire, you must first know what the purpose of the questionnaire is. Because around this goal, every question below it has an extremely clear correlation. Secondly, count the specific figures of various answers to each question. These figures directly reflect the behavior and psychological state of the respondents. Third, it is the most important, complicated and subtle work. What can you know directly or indirectly by analyzing the figures? For example, breakfast for ten people, two people eat at home, five people buy it outside, two people sometimes eat it, and one person doesn't eat it. Figures show that seven out of ten people have the habit of eating breakfast, and three out of ten people are irregular. If we continue to dig deeper, we will find that most people pay more attention to breakfast, while a few people ignore its function. If the question is more detailed, we can continue to study whether breakfast is scientific among people who value breakfast. It should be emphasized that the results of each item are closely related to the purpose of the survey and are analyzed around the purpose. Fourth, according to the analysis results, make a general conclusion.
My humble opinion is for reference only.
Question 9: How to process the questionnaire data for statistical analysis? You mentioned the statistical analysis table, which is wrong.
There is no such thing.
You can first design the research purpose, make research hypotheses, then make an analysis according to the hypotheses, and then make a table.
I often help others do statistical analysis of this data.