Short Note on Regression Analysis Regression analysis is one of the most extensively utilized method between the analytical models of association employed in business research. Regression analysis tries to analyze the connection between a dependent variable and a group of independent variables (one or more).

One example is, in demand analysis, demand is versely linked to price for normal commodities. We can write D = A – BP, where D is, the demand which is the dependent variable, P is the unit price of the commodity, an independent variable. It is an example of a simple linear regression equation. The multiple linear regressions model is the prototype of single criterion multiple predictor association model where we wish to research the combined impact of several independent variables upon one dependent variable. In the above example if P is the consumer price index, and Q is the index of industrial production, we might manage to research demand as a function of 2 independent variables P and Q and write D = A – BP + C Q as a multiple linear regression model.

Regression analysis is commonly employed for prediction and forecasting, where its use has considerable overlap with the field of machine learning. It is also utilized to understand which among the independent variables are related to the dependent variable, and to take a look at the types of these relationships. In restricted situations, regression analysis enables you to infer causal relationships between the independent and dependent variables. However this can result in illusions or false relationships, so caution is advisable; for instance, correlation doesn’t mean causation.

Objectives of Regression Analysis

– To research a pattern linking the dependent variable and independent variables by establishing a functional relationship between the two. In this equation the level of relationship comes from which is a matter of interest to the researcher in his study.

– To make use of the well-established regression equation for problems concerning forecasting.

– To analyze how much of the variation in the dependent variable is described by the group of independent variables. This would allow him to get rid of particular unwanted variables from the system.

For instance, if 85% of variation in demand in a research can be stated by price and consumer rating index, the researcher may drop additional factors such as industrial production, extent of imports, substitution effect etc. that may add only 15% of variation in demand provided all the causal variables are linearly independent.