A further validation of the state-trait distinction in anxiety research