Adjusting for cross-cultural differences in computer-adaptive tests of quality of life