Sei sulla pagina 1di 3

Multi regression model

The simplest possible multiple regression model is three variable regression, with
one dependent variable and two explanatory variables .

For example

yi=b1+b2 x2i+b3x3i+ui

Where y is the dependent variable (regressed) X2 and X3 are the explanatory


variables (regresses) u is the stochastic distribution term and b1 is the intercept
term and B1 is the intercept term which gives the mean or average effect on y of all
the variables excluded from the model. The co-efficient B2 & B3 are called the
practical regression co efficient

Liner model

We assume that the multiple regression model is liner in the parameters, the value
of the regressors are fixed in repeated sampling and that there is sufficient
variability in the values of the regressions interpretation of multi regression
equation

Given the assumption of the classical regression model, it follows that on taking
the conditional expectation of y on both sides of equation , we obtain

E (yi/X2i,X3i)=B1+B2X2i+B3X3i

The above equation gives the conditional means or exception value of the
variables X2 and X3. Therefore as in the two variables case multiple regression
analysis is regression analysis conditional upon the fixed values of the explanatory
variables &it is obtained the average or mean value of y or mean response of y for
the fixed values of the X variables . meaning of partial regression co-efficient.

The meaning of partial regression coefficient is as follows:-

B2 measures the change in the mean value of y E(Y/X2,X3), per unit change in
X2, holding X3constant –it gives respect to X2, holding X3 constant

B3 measures the change in the mean value of Y per unit change in X3, holding
X2 constant.
To assess the true contribution of the X2 the change in Y, control the influence of
X3. Similarly to assess the true contribution of X3 control the influence of X2.

Multiple co-efficient of determination

In the three variable model to know the proportion of the variation in Y explained
by the variables X2 & X3 jointly. The quantity that gives this information is know
as the multiple co-efficient of determination and is denoted by R2.

The value of R2 has between 0 and 1. The higher the grater the percentage of the
variation of Y explained by the regression plane, that is the better the goodness of
fit of the regression plane to the sample observation plane to the sample
observation. The closer R2 to zero the worse the fit.

R2 YX1= b1£YiX1i+b2 Y1X2i

The mean and variance of the parameter estimates B0,B1,B2

The estimates B0,B1,B2 are unbiased estimates of the true parameter of the
relationship between Y,X1 and X2 their mean expected value is the true parameter
itself

E(b0)=b0 E(b1)b1 E(b2)=b2

Tests of significance of the parameter estimates.

Slandered error test

SBq+⎷𝑣𝑎𝑟 (𝐵𝑞)

If SBi > ½ Bi->accept the null hypothesis

If SBi < ½ Bi->reject the null hypothesis

The smaller the slandered errors, the stronger is the evidence that the estimates are
statically significant

The students test of the null hypothesis:-

This is the observed value of the t ratio which will be compared with the
theoretical value of t obtained from the t –table with n-k = n-3degrees of freedom.
If f* falls in the acceptance region, that is t0 .025<t*<t0.025 with n-k, accept the
null hypothesis.

If f* falls in the critical region reject the null hypothesis & accept the alternative
hypothesis -> Bi statically significant.

The grater the value of t*the stronger is the evidence that Bt is significant.

Potrebbero piacerti anche