Different null hypothesis and one sided coefficent tests

[georg_pfleiderer]

Hi all,

For a regression model about inflation hedging I would like to test the zero hypothesis that the coefficient is smaller then one. Does anybody know how to perform single-sided coefficient tests about this.

Does anybody know how I can do that?


Many thanks

Georg

[startz]

Hi all,

For a regression model about inflation hedging I would like to test the zero hypothesis that the coefficient is smaller then one. Does anybody know how to perform single-sided coefficient tests about this.

Does anybody know how I can do that?


Many thanks

Georg

Use EViews to compute the t-statistic in the usual way (the Wald test) just as you would for a two-sided test (c(1)-c(2)=0 or whatever). Then look up the one-sided critical value in a table rather than using the p-value that EViews displays.