Calculateur de Taille d'Effet (Cohen)
Calculez la taille d'effet de Cohen pour votre étude éducative.
Average of all pairwise item correlations
Cronbach's Alpha
0.896
Cronbach's Alpha vs Number of Test Items
Formule
## How Cronbach's Alpha Works Cronbach's Alpha estimates reliability based on the number of items and their average correlation. Higher alpha means the items consistently measure the same construct. ### Standardized Alpha Formula **Alpha = (k x r_avg) / (1 + (k - 1) x r_avg)** Where k is the number of items and r_avg is the average inter-item correlation. ### Interpretation - 0.90+: Excellent reliability - 0.80-0.89: Good reliability - 0.70-0.79: Acceptable reliability - Below 0.70: Questionable to poor
Exemple Résolu
A 20-item survey has an average inter-item correlation of 0.30.
- 01Numerator: 20 x 0.30 = 6.0
- 02Denominator: 1 + (20 - 1) x 0.30 = 1 + 5.7 = 6.7
- 03Alpha = 6.0 / 6.7 = 0.896
- 04Interpretation: Good reliability (0.80-0.89)
Questions Fréquentes
Can Alpha be too high?
Yes. Alpha above 0.95 may indicate redundant items that could be removed to shorten the test without losing reliability.
How many items do I need for good reliability?
More items generally increase alpha. With moderate inter-item correlations (0.2-0.4), 15-20 items usually achieve alpha above 0.80.
What is the difference between Alpha and test-retest reliability?
Alpha measures internal consistency (one administration). Test-retest measures stability over time (two administrations).