Reliability & Validity
Brilliant’s measurement methods are based on and regularly validated through statistical analyses such as factor analysis, regression analysis, and item response theory.
Our measurement methods are firmly rooted in academic theory. In addition, the measures are regularly validated through statistical analysis, including factor analysis, regression analysis, and item response theory models. Brilliant’s database contains more than 1 million records from employee surveys conducted across numerous organisations in various industries.
Our indexes demonstrate strong psychometric properties in terms of reliability and validity.
Reliability
Cronbach's α typically ranges between 0.8 and 0.9 for our various indexes. This indicates that our measurement scales are accurate and display good internal consistency.
Validity
Our measures exhibit strong content validity as well as construct validity (both convergent and divergent). This means that our measurement scales are accurate and assess what they are intended to measure. Furthermore, our results are generalisable, as they produce consistent outcomes across different organisations, reflecting good external validity.
💡 Reliability and validity are also key considerations in the development of tailor-made instruments that we create for our clients.