In this series of Statistical Sidebars, I generally pick one test or analytical technique to look at in some detail to help readers understand why it was used, situations in which it can be used, or how to interpret its results. In this sidebar, I would like to take a bit of a step back and look more broadly at how an author approaches analyzing a set of data. One can think of the conduction of a research study as incorporating two linked components. On the one hand, a researcher must design the study in order to answer some question that is at hand. A related component is planning how the data will be analyzed once they have been gathered. In most research studies, a researcher goes into the study with a plan for both how the data will be gathered and how it will be analyzed. Circumstances might call for adjustments, but, typically, knowing what form the data will be in and how it was obtained will guide how it can be analyzed. In the article entitled, “The Comparison of Physical Fitness, Anthropometric Characteristics, and Visual Acuity of Goalball Players with Their Right Shot Performance,” the authors go into detail about how they collected their data and what the measures were, then they use some inferential statistics to make connections within the data set. One of the first statistical techniques these authors use is a Shapiro–Wilk’s test. This method is a way of checking to make sure that the data conform closely enough to a normal distribution to allow the use of parametric statistical tests (like analysis of variance [ANOVA] and t-test). Authors do not always report that this step was taken; thus, it is sometimes assumed to have been done. Having passed this hurdle, the authors use a series of ANOVAs to determine whether different variables had an effect on right shot performance. Because these other variables were made up of groups (like left hand and right hand grip force or three levels of vertebral motility), the ANOVA is a good test for this step. If any significant differences are found in one of the ANOVA tests, this step would be followed up with a series of t-tests that compare one level of a grouping variable against another to determine exactly where the differences lie. For example, for vertebral motility, if the ANOVA had a significant

中文翻译：

---

复杂分析方法的组织

在这一系列的统计侧边栏中，我通常会选择一种测试或分析技术来详细研究，以帮助读者了解使用它的原因、可以使用它的情况或如何解释其结果。在这个侧边栏中，我想退后一步，更广泛地了解作者如何分析一组数据。人们可以将一项研究的进行视为结合两个相互关联的组成部分。一方面，研究人员必须设计研究以回答手头的一些问题。一个相关的组成部分是计划在收集数据后如何分析数据。在大多数研究中，研究人员在进行研究时会就如何收集数据和如何分析数据制定计划。情况可能需要调整，但通常情况下，了解数据将采用何种形式以及如何获得数据将指导如何对其进行分析。在题为“门球运动员的体能、人体测量学特征和视力与其正确射门表现的比较”的文章中，作者详细介绍了他们如何收集数据以及测量的内容，然后他们使用了一些推论统计数据以在数据集内建立联系。这些作者使用的第一个统计技术之一是 Shapiro-Wilk 检验。此方法是一种检查方法，可确保数据与正态分布足够接近，以允许使用参数统计检验（如方差分析 [ANOVA] 和 t 检验）。作者并不总是报告采取了这一步骤；因此，有时假定已经完成。过了这关，作者使用一系列方差分析来确定不同的变量是否对正确的击球性能产生影响。因为这些其他变量由组（如左手和右手握力或三个椎体运动水平）组成，所以方差分析是这一步的一个很好的测试。如果在其中一个 ANOVA 检验中发现任何显着差异，则此步骤之后将进行一系列 t 检验，将分组变量的一个级别与另一个级别进行比较，以确定差异所在的确切位置。例如，对于脊椎运动，如果方差分析具有显着的 方差分析是这一步的一个很好的测试。如果在其中一个 ANOVA 检验中发现任何显着差异，则此步骤之后将进行一系列 t 检验，将分组变量的一个级别与另一个级别进行比较，以确定差异所在的确切位置。例如，对于脊椎运动，如果方差分析具有显着的 方差分析是这一步的一个很好的测试。如果在其中一个 ANOVA 检验中发现任何显着差异，则此步骤之后将进行一系列 t 检验，将分组变量的一个级别与另一个级别进行比较，以确定差异所在的确切位置。例如，对于脊椎运动，如果方差分析具有显着的