Question

1. For what kind of models can we use OLS?

Answer 1

OLS means Ordinary Least Square. It is the best and most approximate estimation method used for simple linear regression models. If the linear models meets the OLS conditions, such linear models can be said to get best possible estimates. WIth a data, OLS will fit a function closely. For that, the sum of squared errors from the data have to be minimized. Now-a-days scientists and sociologists used to do regression very scarcely, that too using one independable variable. Whereas, OLS performs the same with number of variables.

In Simple linear regression models,

• They help in exploring bivariate as well as multivariate variable relationships. Here, one variable depends on another or combined variables.
• Correlation coefficient figures out two variables are linked with other, however does not reveal the kind of relationship.
• Relation between political science and economics variables are not accurate. Only the definition will be correct.
• This helps to identify the average relationships which could not be understood by just a one-round view of the data. Intense formulation of structural components of hypothesized variables relationship which are random. As an example is the positive relationship existing between unemployment and government spending.