AllFreePapers.com - All Free Papers and Essays for All Students
Search

Classical Linear Regression Model

Autor:   •  July 12, 2015  •  Study Guide  •  7,410 Words (30 Pages)  •  1,214 Views

Page 1 of 30

(Limited Dependent Variable Models)

Classical Linear Regression Model (CLRM)

Requirement: Dependent variable is quantitative within ratio interval

Focus on drivers, regressors, right side, s, dependents[pic 1]

[pic 2]

[pic 3]

        = endogenous, dependent[pic 4]

        = exogenous, independent, fixed value[pic 5]

        = stochastic, error term, random, sablay ng mga s[pic 6][pic 7]

Economic theory will dictate which  to choose[pic 8]

Direction of causality is always towards [pic 9]

        In matrix form:

[pic 10]

[pic 11]

Objectives:

        Find s (problem of estimation)[pic 12]

        Perform inferences on s (problem of inference)[pic 13]

                Individual tests

                         vs         [pic 14][pic 15][pic 16]

                Joint test

                         vs  at least 1  is not zero[pic 17][pic 18][pic 19]

                100 % on  (construction of confidence interval)[pic 20][pic 21]

                                [pic 22][pic 23]

                Goodness of fit

                                if  is 0.9, 90% of variation explains [pic 24][pic 25][pic 26]

                        [pic 27]

                 (population regression function (PRF))[pic 28]

                 (sample regression function (SRF))[pic 29]

 marginal contribution independent of  ceteris paribus[pic 30][pic 31]

for every unit change in ,  will change by  ceteris paribus[pic 32][pic 33][pic 34][pic 35]

Assumptions:

         is MVN (multivariant normal)        error is normally distributed[pic 36]

                error vanishes in the long run[pic 37]

                heteroskedasticity[pic 38]

        Non-correlation

                exogeneity assumption[pic 39]

         has full rank         non-multicollinearity[pic 40][pic 41]

Gauss-Markov Theorem

        If assumptions above are true, OLS is BLUE

Binary Response Models ( is dummy; 0,1 failure-success basis)[pic 42]

        Linear Probability Model (LPM) (OLS)

        Logit Model (Logistical link)

        Probit Model (Standard normal link)

Multinomial Response Model ( is multinomial, 3 or more)[pic 43]

        Multinomial Logit Model

        Multinomial Probit Model

...

Download as:   txt (16.7 Kb)   pdf (400.7 Kb)   docx (21 Kb)  
Continue for 29 more pages »