Why Is Reliability Analysis Essential?
Every study that uses a questionnaire or scale must demonstrate its reliability. Cronbach's Alpha (α) is the most widely used reliability coefficient — it estimates internal consistency, i.e., how well all items measure the same underlying construct. Thesis committees and journal reviewers will always ask for it.
How to Interpret Cronbach's Alpha
- α ≥ 0.90: Excellent reliability
- 0.80 ≤ α < 0.90: Good reliability
- 0.70 ≤ α < 0.80: Acceptable reliability
- 0.60 ≤ α < 0.70: Questionable — use with caution
- α < 0.60: Unacceptable — scale or data needs review
Running Reliability Analysis in SPSS
Go to Analyze → Scale → Reliability Analysis.
- Move all scale items to the Items box.
- Confirm that Model is set to Alpha.
- Click Statistics: check Item, Scale, and Scale if item deleted.
- Click Continue → OK.
Understanding the Output
The Reliability Statistics table shows the overall Cronbach's Alpha. The critical table is Item-Total Statistics:
- Corrected Item-Total Correlation below 0.30 → the item may not fit the scale well.
- Cronbach's Alpha if Item Deleted: if removing an item substantially raises alpha, consider dropping it.
Subscale Reliability
If your scale has multiple subscales (factors), run a separate reliability analysis for each subscale. Report each alpha individually alongside the overall scale alpha.
APA Reporting Example
Internal consistency of the scale was assessed using Cronbach's alpha. The overall scale demonstrated good reliability (α=.87). Subscale alphas ranged from .79 to .84, indicating acceptable to good internal consistency (Nunnally, 1978).
Boss Statistics Support
Unsure whether to drop an item or how to write up your reliability findings? Boss Statistics provides expert guidance on scale analysis and APA-formatted reporting.