reliability of the coefficients in a regression
res=reliability(res,w,p,np)
res = the results of a -univariate- regression
w = a (1 x 4) vector of reliability outcome:
- w(1) = significant only on a sub-sample
- w(2) = significant only on the sub-samples
- w(3) = significant only on the whole sample
- w(4) = significant on the whole sample and on a sub_sample
p = a size level
np = 'noprint' if the user does not want to print the results
res = a tlist with:
- res('meth')='reliability'
- res('namey')= a string, the name of the endogenous variable in the regression
- res('namex')= the string vector of the names of the exogenous variables in the regression
- res('reliab')= the real vector of the reliability of the coefficients
load(GROCERDIR+'data/bdhenderic.dat'); bounds('1964q3','1989q2'); rols=ols('delts(lm1-lp)','delts(lp)','delts(lagts(1,lm1-lp-ly))','rnet', 'lagts(1,lm1-lp-ly)', 'const'); reliability(rols,[0.3 0.6 0.4 0.7],0.05) // From Hendry and Ericsson estimation, calculates the reliability of the coefficients at a 5% level, using Hendry and Krolzig proposed parametrisation (w=[0.3 0.6 0.4 0.7]).