Quantitative data in medical papers statistical analysis of errors in the Analysis and remove doubts.docVIP

 PAGE \* MERGEFORMAT 11 Quantitative data in medical papers statistical analysis of errors in the Analysis and remove doubts [Keywords:] statistics; Medicine; papers; statistical analysis; non-parametric statistics One of the problem In the medical manuscript in the papers or, if the observations are quantitative data, often need to compare different experimental conditions, whether the average difference between the statistical significance, this time, known as the statistical analysis of quantitative data. How can we achieve the proper quantitative data of statistical analysis? The key is two things: First, check whether the quantitative data to meet parametric test a prerequisite; Second, the correct Analysis of quantitative data corresponding to the type of experimental design. The first point can often be achieved through the statistical software, while the second point requires analysis of people with that knowledge, a reasonable choice is possible statistical analysis method. However, as long as the research projects involving two or more factors, the actual workers to correctly determine the type of occasion very little experimental design, therefore, in the medical papers or manuscript, the countless mistakes in this regard. To sum up, no more than the following two types of errors: (1) When the quantitative data does not meet a prerequisite for parametric test (independence, normality and homogeneity of variance), the blind method of applying parametric test (usually for a specific design quantitative data under the t-test or analysis of variance); (2) quantitative data regardless of type of experimental design corresponds to what is applied to all single-factor two-level (or called into the group) designed to test methods of quantitative information (such as t test or rank, and test) or single-factor multi-level design of quantitative data analysis methods (such as multi-level design of single-factor analysis of variance of quantitative informat