Reliability and validity must be tested in the process of questionnaire survey, because questionnaire survey is often only a part of the whole project. Under the correct project objectives, there must be another survey reliability and effective analysis to support the survey results, so that our questionnaire survey can be credible and the results can tend to correct data.

Reliability refers to the consistency, stability and reliability of test results, and internal consistency is generally used to express the reliability of the test. The higher the reliability coefficient, the more consistent, stable and reliable the test results are.

System error has little influence on reliability, because system error always affects the measured value in the same way, so it will not cause inconsistency. On the contrary, random errors may lead to uncertainty, thus reducing reliability. Reliability can be defined as the degree to which random error r affects the measured value. If R=0, it is considered that the measurement is completely reliable and the reliability is the highest.

Generally, if it is a questionnaire containing a scale, it is necessary to analyze the reliability and validity. No-metric questionnaire can be described in written form, no matter what kind of questionnaire, it should be described in the paper to prove the reliable data quality.

If it is a self-made scale, it is generally necessary to pre-test, that is, distribute the questionnaire in a small range, analyze the reliability and validity, and modify or delete the items with low reliability and validity, so that researchers can make some adjustments to the initial questionnaire and form the final version. Of course, formal research still needs to do reliability and validity analysis.

Validity and reliability are two basic conditions of a good measuring tool. The relationship between validity and reliability can be summarized in one sentence: reliability is a necessary condition for validity, not a sufficient condition.

Reliability is a necessary condition for effectiveness, that is to say, an indicator must be reliable if it is effective, and it cannot be correct if it is not credible. However, reliability is not a sufficient condition for validity, that is to say, with reliability, it is not necessarily effective.

Strictly speaking! Not all questionnaires are suitable for reliability and validity analysis, which is mainly aimed at scale questionnaires, but it is not suitable for reliability and validity analysis if it is only some objective reality (such as age, gender, occupation, vehicle, salary, etc.). ) being investigated! To judge whether some variables are suitable for reliability and validity test, we should pay attention to the following points:

(1) Latent variable: directly unobservable variable, which mainly reflects people's cognition and subjective will.

(2) Measurable: Measurable variables are generally ordered or equidistant variables, not classified variables like places.

(3) Equidistant scale between variables: for example, the measurement data obtained by 5-point or 7-point scoring method.