Table 3 Hypotheses to assess construct validity

1. The number of items where people indicate that it influences their dignity (very) much (score 4 or 5) correlates strongly with the single item score for overall personal dignity (on a scale from 1 to 10).

Although even a single symptom or experience may suffice to violate dignity for an individual nursing home resident, we expect that – on study population level – the more items influence dignity to a large extent, the lower the single item score for personal dignity.

2. Both the score for quality of life (on a scale from 1 to 10) and the score on the WHO Well-Being Scale (on a scale from 1 to 100) correlate moderately to strong with the single item score for personal dignity (on a scale from 1 to 10), though the correlation with the WHO Well-Being Scale is stronger than with the score for quality of life.

These expectations arise from the results of our interview study [22, 29], in which we noticed that many nursing home residents associated 'quality of life’ with their physical health status. Because personal dignity encompasses relational aspects as well, we expect 'well-being’ – which might have a more holistic connotation – to be more closely related to the concept of dignity.

3. Both the score on the Karnofsky Performance Status Scale and the score on the Barthel Index correlate low to moderately with the single item score for personal dignity (on a scale from 1 to 10), though the correlation with the Barthel Index is stronger than with the Karnofsky Performance Status Scale.

Whereas the Barthel Index measures physical functioning on 10 Activities of Daily Living – and the Karnofsky Performance Status Scale simply by one question – we expect more variation in the scores on the Barthel Index, and therefore a stronger correlation with personal dignity. However, we expect these low to moderate correlations since physical functioning is only one aspect of personal dignity.