Constructing Groupings by Use of STATISTICA Software Package
STATISTICA software package for statistical analysis incorporates a wide range of advanced statistical methods. Quite often they are preceded by aggregating statistical survey data, which main component is their grouping. Although this phase of statistical data processing is relatively simple, the manual process of aggregation can be time-consuming given the need to process large data arrays, not mentioning a high probability of errors. Therefore, the all-purpose STATISTICA software package is a logical and reasonable tool for grouping of data.
The article shows the grouping algorithm in STATISTICA software package, with focus on setup when constructing tables of frequencies of discrete and continual characters. Various options of grouping are scrutinized, with providing examples of their visualization.
A large number of STATISTICA parameters offers ample opportunities for constructing user tables, but users often are not aware of these options or do not know how they can be applied. Yet, the apparently simple grouping process in STATISTICA software package can sometimes require the knowledge of fine mechanisms for its setup. The article gives a detailed description of the mechanisms for creating interval margins when applying the parameter “approximate number of intervals”.
The standard algorithm for selection is analyzed, allowing a user to limit the number of groups in a grouping. STATISTICA allows for using a number of grouping parameters, enabling to produce more convenient results or filter them. Thus, setting the clicker for label field “Grouping” in the position “Integer Categories” (integer intervals (categories)) initiates the grouping only for integer values of a variable, by excluding the observations containing its fractional values.
When only standard parameters are used, it will be impossible to form uneven or open intervals. This issue is out of focus in specialized literature and Internet sources. The article shows the algorithm for constructing open intervals by user-set conditions and the process of creating these conditions. This option allows for forming both closed and open intervals by solving all the problems in time of grouping. Because creating such conditions is time consuming, they should be preserved if they are required for further use.
Setting up of STATISTICA software with missing data is analyzed. Its application will be advisable when a grouping for two or more variables is constructed. In this case, a separate sheet with a grouping is to be created in the worksheet for each variable.
Fetisov, V. S. (2018). Paket statystychnoho analizu danykh STATISTICA. [Package of statistical data analysis STATISTICA]. Nizhyn: Vydavnytstvo NDU imeni Mykoly Hoholia [in Ukrainian].
Abstract views: 108 PDF Downloads: 101