This article points out that in the EPR experiment, there is no causal connection between the two sides. If a causal model wants to comply with the principle of common cause, it will lead to faster-than-light information propagation. Even if non-local latent variables are used as common factors, it does not mean that the measurement results on both sides can influence each other. They are only related to each other, but not corresponding causal relationships.The "statistical correlation" in the EPR experiment does not have a controllable effect, because the results measured on either side are completely random. The experimenter cannot control the other side to produce the results he wants to transmit by manipulating the instrument. The two experimenters can only discover the correlation between each other's data when comparing them afterwards, and they cannot use this correlation to transmit information, otherwise the incommunicability theorem will be violated.This article draws a causal model diagram to analyze the EPR experiment, and uses the Everett interpretation and the quantum circuit diagram to demonstrate. It shows that the causal paradox in the EPR experiment is actually caused by a misunderstanding of "measurement" and "collapse" in the Copenhagen interpretation. If the Everett interpretation without the concept of "collapse" is adopted, it can be avoided when interpreting the EPR experiment. Paradox of cause and effect.In Everett's interpretation, the physical meaning of "measuring" a particle is to entangle the particle with the measuring instrument. When the local particle is measured, it will only be entangled with the local instrument, so there is no non-local interaction. The state of the particle does not change in any way because the particle on the other side is measured. No matter who measures first on the left or right, there will be no influence on each other. There will be no effect of the result of the first measurement on the result of the later measurement through superluminal action, and there will be no paradox that overturns the order of cause and effect.