Reliability is reliability, which refers to the stability and consistency of the questionnaire survey results when the same object is investigated by the same method, that is, whether the measuring tool (questionnaire or scale) can measure the measured things or variables stably. Specific evaluation methods are:
1, repeated test method. The same questionnaire measures the same group of respondents twice at different times under the same circumstances as much as possible. The interval between two measurements is generally within two to four weeks. The reliability of the scale is evaluated by the correlation analysis between the two measurement results or the significance test of the difference.
2. Staggered method. Two different forms of equivalent questionnaires are used to measure the same group of respondents at different times (usually two to four weeks apart). The correlation between the two measurement results is used to evaluate the reliability of the questionnaire.
3. semi-method. The folding method is to combine the above two questionnaires into one questionnaire (usually the number of questions in the two questionnaires is equal), each as a part, and then investigate the correlation between the measurement results of the two parts.
Secondly, the validity of the questionnaire
Validity refers to the degree to which the questionnaire correctly measures the variables to be measured by researchers. The main indicators and methods for testing validity are: face validity, criterion validity and frame validity.
1, the coupon is valid. Also known as content validity or logical validity, it refers to whether the measured content is suitable for the measurement target or whether the selected item "looks" meets the measurement purpose and requirements. Mainly based on the subjective judgment of survey and design personnel.
2. Validity of standards. Criterion validity refers to whether the data obtained from the scale is meaningful compared with the values of other selected variables (criterion variables). According to the different time span, the validity of the standard can be divided into simultaneous validity and predictive validity.
Homogeneous validity refers to the degree to which other measurement methods of the same characteristic are interrelated. Heterogeneous validity refers to the degree of correlation between the scale and the measurement methods of different characteristics, but it is related to the characteristics in theory. Constructive validity refers to the degree to which theoretical concepts can be measured by measuring tools, that is, if the results obtained by measuring tools are compared with other concepts under the same theory, when there is some expected correlation between them, it means that this measuring tool has a certain degree of constructive validity.
Generally speaking, the reasons for errors in questionnaire survey include: improper measurement content, inappropriate situation and negligence of researchers themselves. The factors from the interviewees may be due to their age, personality, education level, social class and so on. This affected the correctness of their answers. Through the test of reliability and validity, researchers can know whether the questionnaire itself is good or not, which can be used as the basis for revision and avoid making wrong judgments.
The first condition for the formation of problem consciousness is that you should be full of curiosity and strong interest in all phenomena and have strong observation ability. For example, when you go down to the relevant practice place, you observe some phenomena and problems, such as: Why does the teacher seem to be performing in class? Why are teachers interested in it, but there are always some students who are not interested in it? We should actively think about it, ask more why, and combine the relevant education and teaching theories we have learned to ask how to be better.
Secondly, we should emphasize information awareness. Information resources are the basis of forming valuable problems. In addition to vivid perceptual information obtained from actual observation, rich theoretical resources are also the key to the formation of the problem. All research achievements in some fields, including the latest formulation, should be accumulated in the brain, organized according to the optimized structure, and internalized in the heart. For the existing problems, we should learn to think from another angle, be good at exploring new problems that have not been studied or are not clear enough, and record such problems in time.
Then learn to think, including the methods and angles of thinking. Only those who can think and are good at thinking will put forward meaningful questions to the point and give clear ideas for solving problems effectively.
In the process of doing scientific research, if you encounter problems anytime and anywhere and think deeply in time, I believe there will be great gains.
There are two basic sampling methods: one is to consciously select several representative units from the target population as samples for investigation and research. This kind of sampling is called objective sampling or subjective sampling.
Target sampling method is suitable for typical investigation or research. Its characteristic is that the representativeness of each unit as a sample is determined by the judgment of experts related to the research topic. The reliability and accuracy of the results must be judged and evaluated by experts in this field. Using target sampling, we can infer the relevant characteristics of the population, and we can also properly consider the relevant factors involved in the study. It can also make some information of known sample units play a role in sampling. These are the advantages of target sampling. In addition, target sampling is more convenient in data collection. When using target sampling, the probability of each member of the target population being sampled cannot be calculated by objective methods, so the sampling error and accuracy cannot be evaluated by objective methods, which is the biggest disadvantage of target sampling.
The other is to take samples from the members of the target population according to the random principle, which is called random sampling and probability sampling. The biggest advantage of random sampling is that it can objectively calculate the reliability and accuracy of the survey results in a probabilistic way. And the sample size can be determined according to the actual situation and the required accuracy. Therefore, we should fully understand the significance of random sampling in educational science research, especially in large-scale evaluation research.
Simple random sampling (SRS) is the simplest and most basic random sampling. Its characteristic is that the samples are determined by random number table or lottery, thus ensuring the equal extraction probability of all members in the group.
If the number of samples is large, simple random sampling will be difficult, so mechanical random sampling can often be used instead.
The advantage of adopting simple (or mechanical) random sampling is that the sampling error can be estimated objectively according to probability theory.