In psychological research, this situation often occurs: Based on a specific theory, it is speculated that a variable is the intermediary variable between the other two variables, but the expected mediating effect is often not tested. Although there are many reasons for this situation, the direct confusion of researchers is often: is the theory based on wrong? Is the prediction wrong? Or are both wrong?

In fact, it may be another case: the theory is right and the reasoning of prediction is right. However, the nature of the variables investigated is very special, at least not as the researchers originally thought, but more subtle, so that some effects are difficult or even impossible to be investigated. Therefore, it is particularly important to further consider and classify the variables in psychology. Perhaps, such a classification is a way to explain the above phenomenon.

Spielberger, gorschuch, & lushene (1970) divided anxiety into two independent components: state anxiety and trait anxiety, and developed a questionnaire. State anxiety is a temporary emotional response, while trait anxiety is a relatively stable personality tendency. Since then, these two terms and variables appeared in psychological research, which greatly deepened the research on anxiety (Bernstein & eveland, 1982). Thus, it can be seen that putting forward some concepts appropriately and distinguishing some variables scientifically can promote psychological research.

Based on the ideas of Spielberger, gorsch, & lushene (1970), we can distinguish state variables from trait variables in psychological research. State variables are transient reactive variables, such as passion, short-term motivation, specific needs, etc. Trait variables are relatively stable personality tendency variables, such as extraversion, intelligence, etc. In the research, the state variables may change too fast, that is, when the two test tools are implemented, the corresponding state has returned to the original situation; Trait variables are difficult to change. Either way, it will be difficult to get the expected results.

Since the change process of some state variables may be too fast, it is meaningless to study such variables? On the contrary, such state variables may be extremely important, and it is of unique significance and value to study them. In reality, people often say, "I just did it in a flash. It's too late to repent Such a "flash" is abstracted as a psychological variable, which is a typical state variable. It can be inferred that such state variables play a key role in individual decision-making, especially in major decisions.

On the other hand, the concept of catalyst in chemical research also suggests that the above state variables may be proposed and studied. The catalyst only affects the reaction rate (generally accelerating), but does not affect the energy change before and after the chemical reaction. In other words, the catalyst has no change before and after the study of the reaction. The mediating variables in psychological research may also have this situation, that is, one or some variables actually affect the individual's psychological process, but the final result does not appear to have the effect of this or these variables. The reason for this is that psychological research often does not observe the reaction process.

This method can guide the research work. If the intermediary variables selected by students are state variables that change too fast, they may not get the expected results. In the students' opening report, it is often seen that the variables to be manipulated by students are in fact idiosyncratic, which is inappropriate. If students can consider from the perspective of trait variables and state variables, there will be no such problem.

Obviously, there are many types of variables in psychology, especially according to the situation, some variables can be viewed from different perspectives, so the situation of variables is more complex. Compared with physical variables and physiological variables, psychological variables can be divided into trait variables and state variables in psychological research. Especially, in the analysis of mediating effect, we need to consider the change rate of state variables as mediating variables. The state of some variables is very short. In the research, for example, in the process of psychological measurement, it is often necessary to test multiple questionnaires. Therefore, when testing another questionnaire, the state variables of the subjects have returned to the original situation, or there are other changes due to the influence of the current testing tools. Therefore, it is impossible to test the expected changes.