Are self-experienced neuropsychological deficits able to indicate a liability to schizophrenia? Results of a comparative study of 1st-degree relatives of schizophrenics