Misspecification

 What is Misspecification?

Misspecification is part of the econometrics.One of the most common specification error is to estimate and equation that emits one or more influential explanatory variables or an equation that contains explanatory  variables that do not belong to the true specification.

Consequence of omitting influential variables explanatory variables that play an important role in the determination of the dependent variable causes these variables to become a part of error in the population function. Therefore one or more of the CLRM gemsons will be violated. Now we are going to explain this in detail considered the population regression function;
Non influential variable
Omission and inclusion of relevant and irrelevant variables at the same time

Y=B1+B2X2+B3X3+u

where B2 not equal to zero and B3 not equal to zero and assume that this is the correct from of this relationship.

If an estimated equation includes variable that are not influenced shall the problem is less serious so in this case is that the correct equation is

Y=B1+B2X2+u

And this time estimate

Y=B1+B2X2+B3X3+u

Where X3 is wrongly included in the model specification.

In this case suppose the correct equation is

Y=B1+B2X2+B3X3+u

And we estimate

Y=B1+B2X2+B2X2+u*

hare we not only omit the relevant variable X3 but also include the non influential variable X4 at the same time. This was analysed a verb the consequence of the first case are to have biased and inconsistent estimates and II gives in in efficient estimates.

Various functional forms

Linear

Linear log

Reciprocal

Quadratic

Interaction

Log-linear

Log reciprocal

Log quadratic

Double log

Logistic

Tests for misspecification

Normality of residual

Ramsey RESET test for general misspecification

Test for non nested models.