Internal consistency is based on correlations between different items on the same test.

• A. Agreement among different raters
• B. Correlations among items on the test
• C. The stability of measurements over time
• D. The relationship between the test and external criteria

Answer: (B)

Internal consistency looks at whether the items on a single test hang together as a group, meaning they all tap the same underlying construct. This is assessed by examining how the items relate to each other, specifically through correlations among the items. When the items are measuring the same thing, people’s responses tend to rise and fall together across items, which Cronbach’s alpha uses to quantify overall consistency. That is why correlations among the test’s items are the basis for internal consistency.

Other concepts describe different kinds of reliability or validity. Agreement among different raters refers to inter-rater reliability, not how the items relate to each other. Stability over time refers to test-retest reliability, which looks at consistency across occasions. The relationship between the test and external criteria pertains to validity, specifically criterion-related validity.