Hetroscedasticity

Lecture 16. Heteroskedasticity In the CLR model Yi = ?1 X i1 + ? 2 X i 2 + + ? K X iK + ui , i = 1, , n hotshot of the supposals was Assumption 3 (Homoskedasticity) All ui s digest the equivalent variance, i.e. for i = 1, , n Var (ui ) = E (ui2 ) = ? 2 When is this a bad assumption? If omitted variables are non correlated with the included variables (assumption 1), but have a disparate order of magnitude for (groups of) observations. Cross-sectional selective information on units of assorted size, e.g. states, cities. Omitted variables may be larger for more than inhabited states or cities. Cross-sectional data on units at different points in time. Omitted variables may be more outstanding at some points in time. Cross-sectional data on units that daring different restrictions on their behavior. For instance, high income individuals have more manners in their spending. Example of second case: likeness amongst income and be intimate. Data on 222 univ ersity professors for 7 schools (UC Berkeley, UCLA, UCSD, Illinois, Stanford, Michigan, Virginia) See graphs broadside mutation in income (in 1000$) increases with spirt experience Variation in sexual relation income first increases and then decreases Is consistent because income is higher(prenominal) if more work experience 200 Salary (1000$) and work experience ( geezerhood since Ph.D.

) cl SALARY 100 50 0 0 10 20 30 40 50 age 5.5 Log(Salary) and work experience (years since Ph.D.) 5.0 LNSALARY 4.5 4.0 3.5 0 10 20 30 40 50 YEARS Note that logarithm slip reduces cha nce variable in income with experience. Why?! If variation in income increases proportionally with income level, then variation in relative income does non change with income level. Example Income with work experience 4 years: 30,40,60 with direct difference 10, 30, relative difference 33%,100% and log difference 0.29, 0.69 (all relative to lowest) Income at work experience 8 years: 90,120, 180 with absolute difference 30, 90, relative difference...If you fatality to get a safe essay, order it on our website: BestEssayCheap.com

If you want to get a full essay, visit our page: cheap essay