2 x K contingency tables having both ordinal and nominal categories are often encountered in various types of studies. Such data are referred to as 'mixed' categorical data in this article. To apply a method for ordered categorical data one has to discard the nominal categories, and to apply a method for nominal categories one has to discard the ordering information inherent in the ordered categories. Therefore, investigators often either discard observations in nominal categories or discard the ordering of the categories before analysing such data. Some information will be lost in both approaches. A method for analysing data in 2 x K 'mixed' tables is proposed in this paper which can be considered as an extension of well known methods for nominal and ordered categories. The proposed method utilizes observations in the nominal categories as well as the ordering information. If all the categories were ordered then the proposed method reduces to the trend test, and if all the categories were nominal then the proposed method reduced to Pearson's chi-square test.
展开▼
机译:在各种类型的研究中经常会遇到同时具有序数类别和名义类别的2 x K列联表。在本文中,此类数据称为“混合”类别数据。为了对有序的分类数据应用一种方法,必须丢弃标称类别,而对于有名的分类应用方法,必须丢弃有序类别中固有的排序信息。因此,研究人员通常在分析此类数据之前要么丢弃名义类别中的观察结果,要么丢弃类别的顺序。两种方法都会丢失一些信息。本文提出了一种用于分析2 x K“混合”表中数据的方法,可以将其视为公称和有序类别的众所周知方法的扩展。所提出的方法利用名义类别中的观察结果以及订购信息。如果所有类别均已排序,则所提出的方法归结为趋势检验,如果所有类别均是名义检验,则所提出的方法归结为Pearson卡方检验。
展开▼