N represents the number of samples. An experiment always needs a control experiment (experimental group and control group). Control experiment (i.e. control experiment) is just a group of experiments with different conditions (i.e. factors) and the same other conditions (factors). Based on the relationship between conditions (factors) and results, the experimental group used for comparison is generally called the control group. It is not difficult to find that the control group plays the role of foil, which is actually equivalent to the role of reference in physics. Since the control group as a foil already has a perceptual or theoretical understanding of the experimental process, results and phenomena, that is to say, the experimenter already knows the results of the control group before actually doing the experiment. If it is found that the experimental results of the control group are inconsistent with the previous results after the actual experiment, it means that the experimental operation process is wrong, and the experiment is declared to have failed and must be redone; If the results of all the object groups in the experimental design are not clear to the experimenter, that is, all the object groups are control experiments, we can judge the experimental group and the control group accordingly. For example, in the experimental design of example 1, the experimenter has a certain perceptual knowledge of the experimental results of the exposure group (the experimental results are clear), the exposure group is the control group, and the shading group is the experimental group. Similarly, the experimental group and control group in case 2 and case 3 have the same answers. This concept of deterministic control experiment conforms to the cognitive law of verifying or exploring the experimental process, and can be integrated with the method of collecting and analyzing data and information in the process of scientific inquiry, so it can be applied to all experiments.