compare the significant

[Mareah]

Hi all

my question is
If I have 2 independent variables INVAR1 and INVAR2
and the result was significant at 1% for INVAR1 but significant at 5% (or 10%) for INVAR2, can I say that INVAR1 affect more on the dependent var than INVAR2 ?

(note: INVAR1 and INVAR2 were test on different equations but they are a proxy for one thing and other variable in equation is the same)
e.g INVAR1 is long-term debt and INVAR2 is short-term debt
dependent var is performance

so can I say: long-term debt affect on performance more than short-term debt?

[Lazar]

If i have understood you correctly no. Just because relationship A is significantly different from zero but relationship B is not does not mean relationship A is significantly different from relationship B. You will need to test this using something like a Wald test.

[Mareah]

umm
well let me explain more:
if the first model
var = INVAR1 + other variables
var = INVAR2 + other variables

the result for INVAR1 is significant at level 1%
and the result for INVAR2 is significant at level 5% or 10%
can I say that INVAR1 affect on var more than INVAR2 ?

thanks

[Lazar]

No for the reason I posted above.

[Mareah]

ok

thanks a million