Help please...Hausman test
Posted: Thu Sep 23, 2010 3:15 am
Hi all
i am currently using random effect model for survey panel data. if i test for consistency of random effect by hausman test. For many variables model gives me NA values with a message "* Cross-section test variance is invalid. Hausman statistic set to zero." below Hausman statistics (prob.=1).
with this message and many NA in final cross section random effect model (ruduced rank equation), what can I infer about the results are results valid or not? am i doing any mistake ?....
confused...
i am currently using random effect model for survey panel data. if i test for consistency of random effect by hausman test. For many variables model gives me NA values with a message "* Cross-section test variance is invalid. Hausman statistic set to zero." below Hausman statistics (prob.=1).
with this message and many NA in final cross section random effect model (ruduced rank equation), what can I infer about the results are results valid or not? am i doing any mistake ?....
confused...