If there are many phenomena or research problems involving many variables that should be analyzed simultaneously, then why in practice a lot of information data processing involves only one or two variables only? This is because a lot of complexity in multivariate statistical methods, which do not allow the calculation manually.
If there are many phenomena or research problems involving many variables that should be analyzed simultaneously, then why in practice a lot of information data processing involves only one or two variables only? This is because a lot of complexity in multivariate statistical methods, which do not allow the calculation manually. For example, a simple regression methods and multiple regression; simple regression involving only one dependent variable and one independent variable, often used in practice because the search process is quite simple regression coefficients. By adding an independent variable, called the multiple regression, the counting process becomes complex. If the independent variables into three or more, the ability of manual calculation becomes almost impossible.
Similarly to many other multivariate methods, such as factor analysis, discriminant analysis, the level of complexity may exceed the multiple regression method. But the development of information technology in the 1990’s, particularly in the field of statistical software, allowing the calculation of the multivariate method is fast and precise. Software such as SPSS, Eviews, Minitab or SAS can process data in a number of multivariate tens or even hundreds with satisfactory results. Even now the statistical software development, particularly SPSS, has penetrated the activities of data mining and neural network, which is a ‘continuation’ and multivariate, with higher levels of complexity.
In practice, almost all multivariate methods, unless multiple regression involving only two independent variables, it may not be resolved by manual calculation. All of them rely on information technology (statistical software) for processing; what I can not imagine the statisticians because of human limitations, it becomes possible and easy in today’s information age. Because multivariate methods involves many variables, then the calculation becomes much more complex than the analysis using only one or two variables. In this case, although there are various theories, such as the theory of factor, discriminant, and so the theory, in practice such theories would be very difficult to apply without the aid of a computer.