When to use this
Use this when several indicators are intended to measure the same reflective construct.
Why it matters
Reliability concerns score consistency; it is not proof of validity or unidimensionality.
Step-by-step instructions
1. Confirm the construct and item direction.
2. Open Statistical Analysis and choose Reliability.
3. Select the approved construct or item set.
4. Review missing-data handling.
5. Run the analysis.
6. Inspect alpha, item diagnostics, and any supported omega evidence alongside the measurement model.
Expected outputs
Reliability coefficients, item-level diagnostics, effective sample size, and analysis provenance.
How to interpret the results
Interpret coefficients in context, including item count, construct breadth, and measurement assumptions. Avoid universal pass/fail thresholds.
Common mistakes
Do not apply reflective internal-consistency rules to formative indicators. Do not delete items mechanically to raise alpha.
Was this article helpful?
Your response helps the methodology team improve this documentation.