Statistics is an important means of collecting data and understanding the laws of things, and the occurrence of statistical errors will inevitably interfere with our cognitive work. How to reduce the probability of errors and reasonably avoid these common statistical errors is a necessary prerequisite for improving statistical work. According to the statistical knowledge we have learned, it is often believed that statistical errors are often caused by P-values and statistical significance tests. But the author of this paper did not think so, and put forward new ideas. The author believes that the statistical errors that most threaten the validity of research have little to do with P-values and statistical significance. The authors attribute the sources of statistical errors to improperly designed research designs, unclear research questions, poor data processing, lack of statistical thinking and digital literacy. The author believes that the impact of these problems on statistical errors is much greater than the P value.
正在翻译中..