General Principles of Data Analysis

 

 

1

 

Introduction

The choice of an appropriate statistical technique is a complex issue. Real-life data often contain mixtures of different types of data, which makes the choice of analysis technique somewhat arbitrary. It is quite possible that two statisticians confronted with the same data set may select different methods of data analysis, depending upon what assumptions they are willing to take into account while interpreting the results of analysis. Suppose, there is one dependent variable measured on the interval scale, and five independent variables, of which three are interval-scaled variables, one nominal variable and one ordinal variable with five modalities. In such a situation, some statisticians would use multiple regression analysis, treating one ordinal variable as interval-scale variable and use dummy variables for the nominal variable. Some statisticians may categorize all the interval scale variables and perform an analysis of variance.

However, certain general principles for choosing a statistical technique can be discussed. Besides, certain extraneous factors like the availability of software and its limitations, and availability of time and financial resources, the choice of a statistical technique depends essentially upon the following factors:

(i)   Characteristics of the analysis question;
(ii)  Characteristics of the data;
(iii) Characteristics of the sampling design.