In practically every research project, an initial examination of the data set usually includes frequency tables. Ken Yale D.D.S., J.D., in Handbook of Statistical Analysis and Data Mining Applications (Second Edition), 2018 Frequency Tables Computer packages can be used to categorize continuous variables (recoding) and to tabulate the data in one- or two-way tables (see Program Note 3.1 on the website). There is no need to subdivide the overweight category, even though one-half of observations are in this category. For example, in Table 3.4 the number of obesity levels can be reduced to 3 by combining the underweight and normal categories.
When we encounter categories with no or few observations, we can reduce the number of categories by combining or collapsing these categories into the adjacent categories. If we use the data to guide us, it is easy to obtain apparent differences that are not real but only artifacts. We should use our knowledge of the subject matter, and not use the data, in determining the groupings. In forming groups from continuous variables, we should not allow the data to guide us.
0 kommentar(er)
