Reliability and validity
When we develop questionnaires, it is important to ensure that the data collected is reliable and valid. Reliability refers to the extent to which a measure is free from random error and therefore provides consistent results. Validity indicates how well a question measures what it is specifically designed to assess.
There are different sources of errors and corresponding estimates of reliability as presented in Table 9.2:
Source of error |
What are we asking? |
Estimate of reliability |
How to calculate |
Time-specific |
Is there stability in customer satisfaction over time |
Test-retest |
Correlation between the same survey on different occasions |
Item-specific |
Can the items generalize to... |