⏳ Loading example…
💻
SPSS Guide
How to Read SPSS Reliability (Cronbach's Alpha) Output
Learn to read SPSS's Reliability Statistics and Item-Total Statistics tables together to spot a weak item that is dragging Cronbach's alpha down.
📊 α = .812 (5 items) → .85 if weak item dropped
Step-by-step solution
💻 39 · How to Read SPSS Reliability (Cronbach's Alpha) Output
Interpreting the SPSS Reliability Statistics and Item-Total Statistics Tables
SPSS Guide
Research Question:
A 5-item Employee Engagement scale was run through SPSS's Analyze → Scale → Reliability Analysis. Beyond the headline Cronbach's alpha, the Item-Total Statistics table can reveal a weak item dragging the scale down — how do you spot it?
| Item | Corrected Item-Total Correlation | Cronbach's α if Item Deleted |
|---|---|---|
| Q1 | .68 | .77 |
| Q2 | .71 | .76 |
| Q3 | .24 | .85 |
| Q4 | .65 | .78 |
| Q5 | .70 | .76 |
Reliability Statistics: Cronbach's Alpha = .812, N of Items = 5
-
1Step 1 — Read the headline alpha first: α = .812 for all 5 items — 'good' reliability by George & Mallery's common guideline (.70 acceptable, .80 good, .90 excellent).
-
2Step 2 — Scan 'Corrected Item-Total Correlation' for weak items (rule of thumb: should be ≥ .30): Q3 = .24 is below the .30 threshold — it correlates weakly with the rest of the scale and is a candidate for revision or removal.
-
3Step 3 — Confirm with 'Cronbach's Alpha if Item Deleted': Q3's α-if-deleted (.85) is HIGHER than the overall α (.812) — removing Q3 would improve, not hurt, reliability. No other item's α-if-deleted exceeds .812, so only Q3 is a concern.
-
4Step 4 — Decide and report both versions: Reasonable options: (a) keep all 5 items and report α = .812 (still 'good'), or (b) drop Q3 and report the revised 4-item α = .85 going forward. Either way, disclose which you chose and why.
.812
α (5 items)
.24
Q3 corr. (weak)
.85
α if Q3 deleted
🟢 Good reliability overall (α = .812), but Q3's weak item-total correlation (.24) means dropping it would raise α to .85.
APA-7
The 5-item Employee Engagement scale showed good internal consistency, Cronbach's α = .812. Item-total statistics indicated that Q3 had a weak corrected item-total correlation (r = .24); removing it would have increased reliability to α = .85, suggesting Q3 be revised or dropped in future administrations.