For each ASVAB subtest, the equation shown in Figure 6 was used to compute EAP ability estimates The equation shown in Figure 5 was then used to compute PSDs (using the EAP ability estimates, and assuming a N(0,1) population distribution). The average of the squared PSDs was then computed over applicants and substituted into the equation shown in Figure 4 to compute subtest reliability.
For AFQT scores, reliability was computed using the methodology for computing composite reliabilities reported in Gulliksen (1987; pg. 346-347, Equation 74).
For the CAT-ASVAB, the sample used to calculate the reliabilities consisted of applicants from the 2019 fiscal year (FY2019: October 1, 2018–September 30, 2019). For the P&P-ASVAB, the sample consisted of applicants from FY2011 through FY2019. This large range of years was used because the number of applicants taking the P&P-ASVAB has dramatically decreased over the last decade. Thus, to have a sample size similar to that of the CAT-ASVAB, 9 years’ worth of P&P-ASVAB data were required.
Reliability estimates were computed over all applicants, and by gender (Male, Female), ethnic group (Hispanic, Non-Hispanic), and race (American-Indian/Alaska Native, Asian, Black/African-American, Native Hawaiian/other Pacific Islander, White/Caucasian).
The sample sizes used to compute the reliability estimates across subtests and AFQT scores are given in the table below.