Scribd
Upload
24 views2 pages

Probit Model PDF

The document discusses the probit model, which is a binary response model where the dependent variable takes values of 0 or 1. It defines a latent variable that is related to the dependent variable through a normal distribution. The likelihood function and log likelihood function for the probit model are presented, which are maximized to obtain maximum likelihood estimates of the parameters. It notes that the estimated parameters cannot be directly interpreted as marginal effects.

Uploaded by

sam
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
24 views2 pages

Probit Model PDF

The document discusses the probit model, which is a binary response model where the dependent variable takes values of 0 or 1. It defines a latent variable that is related to the dependent variable through a normal distribution. The likelihood function and log likelihood function for the probit model are presented, which are maximized to obtain maximum likelihood estimates of the parameters. It notes that the estimated parameters cannot be directly interpreted as marginal effects.

Uploaded by

sam
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Tushar Nandi

Probit Model

Dependent variable takes two values, 0 or 1 (Binary response model)

Let us assume that there is a latent variable defined by the regression relationship

is related to y as follows,

With the latent variable specification we have,

Where, F(.) is the distribution function of .

Assume that follows normal distribution.

As long as there is an intercept term in

we can assume zero mean for , i.e.

With this assumption it is possible to estimate only , and not and separately.

A customary normalisation is to assume .


So, we assume

The normality assumption is at the heart of Probit Model.

With these probabilities we can form the likelihood function,

L=

Log likelihood function,

LL =

Maximising this function with respect to we get maximum likelihood estimate of .

Note that, =
Tushar Nandi

But, = = , where, is the normal density function.


The estimated parameter cannot be interpreted as marginal effect. Marginal effect is not
constant; it depends on the value of explanatory variables.

You might also like