Testing and validation
The multi-item well-being scales in NPC's Well-being Measure
have undergone a thorough processes of testing, to ensure that they
are reliable, valid and sensitive to change.
All of the scales have undergone tests to determine the
- Internal consistency - a measure of correlation
between different items in the same scale. It shows the extent to
which individual questions in a scale are measuring aspects of the
same underlying factor. Each scale achieves a score on Cronbach's
Alpha of above 0.7.
- Test-retest reliability - examines how stable results
gathered using the scales are over time.
- Sensitivity to change - tested by comparing mean
scores using a paired samples t-test. Differences in mean scores
are detected at 95% significance.
- Predictive validity - the extent to which a score on a
scale predicts a respondent's score on other measures of
- Face validity - a 'common sense' test of whether the
scales appear to be a good measure, whether questions are
appropriate and whether it behaves in way you might expect.
This process is described in detail in our report Feelings
count, which you can download here.
You can read a shorter paper on the process of validating the
You can see who was in involved in the process of developing the
Well-being Measure here.
You can see all the well-being questions included in NPC's
Well-being Measure by downloading the sample survey here.