Missing values (3%) in the questionnaires were corrected by replacing them with mean values per construct.
Out of 13 completed responses (completed or reconstructed questionnaires), 4 questionnaires (31%) of the addressees marked that all coding in the questionnaire was correct, and 9 questionnaires (69%) that some coding in the questionnaire was incorrect.
A subgroup of 20 DM2 patients pilot tested a draft version of a comparable DCE as used for our study, before it was finished and disseminated among the entire sample, to test whether the wording used in the questionnaire was correct and whether the target population understood the choice tasks.
Any implausibility within the questionnaires caused by the respondent was corrected by standard procedures (e.g. The questionnaire asked for duration in days and the patient answered in years. We calculated the days).
For the assisted-living clients questionnaire indicator scores were corrected for age, education, length of care, help with completing the questionnaire, and kind of care (cleaning house, personal care or assistant) [ 26].
The indicator scores of the mail questionnaire to representatives were corrected for the kind of representative (e.g. spouse, son or daughter), residents' age and education, and length of stay.
Within-person correlation in the reference instrument will reduce the correlation with the evaluated questionnaires and should be corrected for in validation studies [ 50].
The researchers checked all completed questionnaires daily, and mistakes were corrected immediately.
Seventeen of these pairs (2.5%) were misclassified by the questionnaire data and zygosity classifications were corrected.
Questionnaires that did not meet the standards were corrected in the field.
Inconsistencies found in the questionnaire data were checked against the original questionnaires and where data entry errors were found these were corrected.
