Score reliability for tests constructed for African-American populations

Area of Science:

• Psychometrics
• Psychological Assessment
• African-American Psychology

Background:

• Reliability is crucial for valid psychological assessments.
• Previous research has highlighted disparities in assessment tools for minority populations.
• Understanding the psychometric properties of tests used with African-Americans is essential.

Purpose of the Study:

• To systematically review and report test score reliabilities.
• To document the sample sizes used in establishing these reliabilities for tests targeting African-American populations.
• To provide a benchmark for the psychometric quality of existing assessment instruments.

Main Methods:

• A review of existing literature was conducted.
• Data on internal consistency reliability coefficients were extracted.
• Sample sizes (N) associated with each reliability estimate were recorded.

Main Results:

• The average internal consistency reliability was .74 (SD = .16), with a median of .77.
• The median sample size (N) used was 131.
• Tests for individual assessment showed a mean internal consistency reliability of .72 with a median N of 96.

Conclusions:

• The reported reliabilities suggest moderate, but potentially insufficient, psychometric properties for some tests.
• The sample sizes used are often smaller than ideal for establishing robust reliability estimates.
• Further research and test development are needed to enhance the reliability and validity of assessments for African-American individuals.